Updated February 24, 2016 at 10:40 pm: A map showing the various options for a Pape-Queen routing to downtown has been added at the end of this article.
The City of Toronto Planning Division has released the detailed technical scoring for evaluation of six possible routes for a Relief Line linking the core area to the Danforth subway.
The possible routes that survived a preliminary cull were:
- A: Broadview to Queen crossing the Don River at Queen
- B1: Pape to Queen crossing the Don River at Queen
- B2: Pape to Queen, dodging south to serve the Unilever site before crossing the Don and swinging back to Queen Street
- C: Broadview to King crossing the Don River at Queen
- D1: Pape to King crossing the Don River at Queen
- D2: Pape to King crossing the Don River at the Unilever site
Of these, the two Broadview options don’t fare very well, and the real debate is between the four remaining Pape options.
The scoring extends through many categories and pages, but to simplify things and understand just how each factor was ranked, I have consolidated the scores into a spreadsheet. (For the detailed evaluation, refer to the scoring document linked above.)
ReliefLineAlignmentScoringSummary
In place of the scoring system used in the city summary, I have converted each of the symbols to a numeric value:
- Full moon = 4 points
- 3/4 moon = 3 points
- Half moon = 2 points
- 1/4 moon = 1 point
- New moon = 0 points
Within each major grouping of scores, the values were totalled and then normalized to the range of zero to one. For example, if a group contained 4 topics, then the maximum possible score is 16, and each alignment’s score is divided 16 to get a normalized value. This allows the scores from each group to be compared with each other.
At the end, the scores can either be summer individually (each topic counts equally), or the group averages can be totalled (each group count equally regardless of how many topics it contains) and normalized.
On a grand total basis, both the city’s technical evaluation and public feedback came to the same conclusion: even though there are variations between the four alignments, they all average out to a 3/4 score.
The point scores also come out close to the same with higher values on one item balancing lower values on others for, overall, an even ranking.
Quite simply, all other things being equal, the four alignments are more or less equivalent using this scoring system. However, the City opted for alignment B1 because, among other things, is is claimed to be cheaper and simpler to build.
The costs estimates in the City’s evaluation [at p.19] are:
- $3.7-billion for either of the Pape via Queen options regardless of whether they follow Queen or King to the core
- $4.0-billion for the Pape-Unilever-King option
- $4.1-billion for the Pape-Unilever-Queen option
The ridership estimates are generally higher for the via King options, and higher still for the routes serving the Unilever site. In all cases, off-peak riding tends to be low compared to the rest of the network with roughly 2/3 of all boardings on the Relief Line being during the AM or PM peak. The effect is even more striking for configuration where frequent SmartTrack service at 12 trains/hour competes with the Relief Line for customers. This demonstrates the problem of a line aimed at primarily peak period “relief”. As and when we see projections for longer versions of the RL (to Eglinton or Sheppard), the balance of off-peak travel may change because the route will serve new subway territory, not simply provide an alternate route for existing trips.
Despite the even scoring overall, the preferred corridor was selected based on giving some criteria additional weight as shown in the table below.
This evaluation, like so much other recent work, depends very much on the presumed presence of frequent SmartTrack service in the rail corridor. If that is found to be impractical, then the relative importance of the RL changes along with its appropriate alignment to serve the core. This is particularly critical at the Unilever site which would have only SmartTrack serving it if the RL stays on Queen Street.
One point above misrepresents the technical evaluation: both of the Queen Street crossing routes have the same cost estimate regardless of which route they take to the core area (Queen or King) and they share the same river crossing characteristics.
In summary, the choice “makes sense” in the limited context that a frequent SmartTrack service will actually be feasible and will be built. If SmartTrack cannot be provided on a five minute headway with a low fare, then the entire planning process will require a major rethink.
Updated February 24, 2016: The following map shows possible alignments for a route from Pape & Danforth to the core area. Notes on the map talk of a track connection at Danforth to provide access to Greenwood Yard. The technical scoring paper also mentions a southerly route under the rail corridor from Gerrard & Pape eastward. While this is longer, it would avoid the complexity of adding linking curves at Danforth.
Steve Munro 
I seem to recall that about a year ago there was a detailed study of Union Station that said that the maximum number of trains the station could accommodate was about thirty per hour, and that only at peak hours. That seemed very low, and maybe was based on the assumption that most trains would terminate/start their journeys at Union, as is currently the case in rush hour. Presumably if trains are running through Union, as SmartTrack envisages, capacity will be higher. Even so, 12 SmartTrack trains an hour in each direction, or 24 in total, will surely put a strain on Union, even before GO starts to ramp up its other RER lines. Am I missing something? If “surface subway” frequency on SmartTrack is not feasible, what does that mean for the DRL projections? I know you’ve posed that question in your article, but I wonder if the city planners are thinking about it.
Steve: There is a combination of constraints affecting the Union Station corridor including the approach track layouts to the east and west, platform capacity, and the signalling system. These interact in various ways such as a small train blocks routes through the approaches for almost the same length of time, but does not have as large an effect on platform capacity. However, small trains also don’t give much line capacity either. Then there is the question of the service pattern and which trains use which tracks.
All that said, I think City planners are desperately avoiding this issue because it has the potential to completely derail the Mayor’s signature plan. Losing the Eglinton West airport spur was a blow, but it was rationalized as the adoption of a better solution. However, if the frequent SmartTrack service does not pan out, the entire scheme is dead in the water.
LikeLiked by 1 person
Personally I think going down Pape to King at the Unilever site the best option (D2). It hits the condos in the in the Distillery, it hits the financial district and the streets are close enough together that it is walking distance from Queen to Front. It even takes pressure off one of the busiest streetcar lines in the city.
It may be 300 million dollars more but it serves more people. Queen is great and all but there is more development along the section of King from Pape to Yonge than there is on Queen. Just because people think it is a good idea to end off at NPS does not mean it is a good idea overall.
Going along King via Unilever is forward thinking, it supports growth and no doubt will spur further development in the port lands.
Broadview is too far west in my opinion to be of much value. The level of useful development along Broadview is not quite there yet. At least with Pape you have a ton of residential development which for the most part survives off Downtown where things are the closest. Not only that but if the port lands are developed people will be grateful for having a stop there.
LikeLike
So it looks like they didn’t follow your advice and go up Donlands.
Steve: Contrary to popular opinion, I do not control all transit decisions in Toronto.
LikeLike
I strongly disagree. Using the current costing figures (3.7B vs 4B) vs various current metrics of ridership projections (33 to 45% higher for the King-Unilever route), the King-Unilever route is far more cost efficient by every measurement. Choosing Queen is management malpractice writ large, ceteris paribus.
Also, the logic on ST frequency is backwards. The more frequent that GO RER/ST is, especially on the western routes, the better a more southerly RL route on King or Wellington is for making connections. “It’s the network, stupid.”
It’s called the ‘relief’ line, but its long term function is really as the complement or mirror of the University/Spadina line. An RL on Queen is orphaned from the networks that the GTA is supposedly about to spend big bucks on. Queen is the wrong choice in all contexts.
Steve: And that’s why I put “makes sense” in quotes. This is a political decision, plain and simple, to avoid problems with SmartTrack and Relief Line co-existence.
LikeLike
I think you’re being generous Steve. I see more than one misrepresentation on the chart.
I’ll return with more thoughts later.
Steve: Yes, there’s more than one. It shows a left hand, right hand problem even within the Relief Line project. Not a good sign.
The really big problem is that the “rainbow plan” depends so much on SmartTrack which is far from a sure thing.
LikeLike
City Hall wants a station at City Hall. Seems like they have a symptom of “me too”. A City hall station would make for poor interchanges with line 1.
Eglinton station is being shifted so it better aligns with line 5. So what doesn’t work at y&e will somehow work at bay and Queen.
Steve: The shift at Eglinton is within the existing structure by extending the platform into what is now the throat of the tail track north of the station. Moving “City Hall” station so that it overlaps the Yonge line is a much more substantial change, and would require building the new station in a more constrained area than at Bay Street.
LikeLike
The question needs to be, and remain, what makes the most sense in terms of ridership, and overall impact on the network long term. If a Queen Street line ties together better with future growth plans, future location of development etc, great. If it is being selected for reasons of City Hall support, not so much. I worry about this being too far north in terms of balance for current vs future load, but well…
LikeLike
Steve, from your comments above and from this blurb from the map:
do I understand correctly that this version of the RL would not directly connect to the YUS?
Steve: By “directly” if you mean as at St. George or Bloor-Yonge, then no. The connection would be via a passageway from Bay to Yonge Street.
LikeLike
It may be 300 million dollars more but it serves more people.
Steve: And $300m is not a lot of money when we are talking about projects of this magnitude.
BTW it does not relieve the King car in the sense that its heaviest segment is west of Yonge, and the RL is to the east.
Understanding the problem needs to be expanded to understand the relationship between the RER and RL. Put simply, as a subway the relief line is best at building growth and scale, RER/Smart Track provides access to the larger community, the better the interaction between the two the greater the economic potential of the projects. By bringing the lines closer together they increase the economic competitiveness of the corridor significantly. The $300 million increase in cost for a more southerly route of the RL is more than offset by the superior function of a future office district that is supported by both projects. We must see net benefit not simply cost.
LikeLike
Some thoughts on stations in the City Hall area. All measurements are via Google Maps, and are meant for discussion, not actual design.
Distance between the middle of the intersection of Yonge & Queen and the middle of the intersection of University & Queen: Approx 620m
Length of TTC subway platforms: Approx 150m. Note these are platforms only, not mezzanines, stairs and walkways.
150m east of Yonge is halfway between Victoria and Bond streets
150m west of Yonge is just west of James St
150m east of University is the east side of York St
150m west of University is just east of St Patrick St.
In this discussion, ‘Osgoode’ refers to an expanded Osgoode station with platforms under Queen in addition to the existing ones under University.
There’s not enough space to fit three stations, (Queen-Yonge, City Hall, Osgoode) even if the two interchange stations were built as ‘T-shaped’ stations with the Queen stations on the outer sides of Yonge & University.
A single station in front of City Hall with walkways connecting to both of the other stations could be built, but it’s a bit of a walk, approx 235m from the end of each platform.
Building the Queen-Yonge & Osgoode stations as ‘T-shaped’ stations with the Queen stations on the inner sides of Yonge & University leaves a gap of 320 metres between the two, about the same as St. Andrew to Osgoode. East end of Osgoode station would be very close to the SW corner of Nathan Philips Sq, west end of Queen-Yonge would be very close to the SE corner of NPSq.
Steve: I suspect that the greater constraint will be the question of how to actually build the new stations. NPS has the advantage of open space north of Queen and a wider than usual roadway. This is almost certainly the prime reason for putting the station there.
LikeLike
Is the potential future elimination of the King car by the DRL with a King alignment a factor in the decision? Clearly a Queen route would not supplant the 501 and there would be duplication with the surface service. This factor may stunt the growth of the DRL west to Roncesvalles along Queen while spur it on along King.
Steve: The King car is a complex route, and it has many segments. Even if a DRL goes west on a more southern route, it is likely to be under Wellington, and there have been proposals to swing south to the rail corridor near Front (although this may no longer be physically possible thanks to dithering on the DRL while all of the gaps in the condos and office developments fill in). Frankly I do not see the DRL going west for a very long time.
LikeLike
What are these plans? Above ground street-running light rail or subways? Property values in the Pape/Danforth corridor are very high. A full subway requires LOTS of demolition and disruption, while light rail would be less so and could incorporate some existing infrastructure on the CN line.
Steve: A full subway straight up the middle of Pape is what is proposed. The anticipated peak load on the RL, especially when it goes north to Eglinton, is higher than LRT can handle.
LikeLike
Visitors to the New York subway know that there are several stations with long walks to connecting lines. Times Square comes to mind, with a long walk to the Eighth Avenue subway. A Bay Station would have the potential to be a really signature station, and would have to be cheaper than building two separate stations. Am I correct in assuming that a relief line only between downtown and Pape Station would have to use Greenwood Yard? Is there enough capacity for extra trainsets? If not at Greenwood Yard where will the trains be serviced and stored?
Steve: The Relief Line would use Greenwood, but the TTC is looking at a large property near Kipling Station for a new west-end yard. This would split BD between Greenwood and “Kipling” leaving lots of space for the RL.
LikeLike
I was at the meeting last night (Feb 25) at Riverdale Collegiate, which was primarily about the DRL. Most people who came out, and the place was packed, seemed to be enthusiastically in favour of a station at Queen and Pape, which can’t happen if the subway follows the Queen St. alignment, as, apparently, stations can’t be built on curves. To accommodate a station there, the subway would have to continue down Pape and then westbound under Eastern. On the other side of the river, it could then follow the Richmond St. route to downtown. This makes way more sense than duplicating service along Queen, and a station on Eastern near the Don would serve both the Unilever lands and the big condo development going in on the former Honda dealership and Tepperman site. A big bonus, to offset the additional cost of crossing under the river at Eastern rather than Queen, would be that way fewer people and businesses would be affected or inconvenienced during construction. Plus its terminus would be at the northern edge of the financial district, which is probably where most riders who’ve transferred from Line 2 at Pape will be headed.
Steve: What is astoundingly annoying about the “preferred” route is that it ignores many factors like this and seems, primarily, to be avoiding conflict with the Mayor’s SmartTrack scheme. the sooner Metrolinx torpedoes the idea of very frequent ST service, the better. We can then get on to planning a network based on what can actually be accomplished rather than left over election literature and consultants whose are long past their sell-by dates.
LikeLike
For talking to serveral city planning and TTC folks at the public meeting (Richview HS) I came away with the impression that the the section from Pape to Gerrard Sq. and the crossing of the Don at Queen were pretty much locked in.
The TTC person said that the Lever site has no location on the lands where enough straight exists to drop in a station, it would have to go further East on Eastern and would need a lot of expropriation to happen. He also said that a GO or ST station faces the same problem (no straight track) and a stop on the Lakeshore East line would have to be east of the Don or north of Queen at Jones. If this is the case, dropping down to the Lever site is a waste of funds. The better solution seems to be a stop at Queen/Broadview with an extension of the Broadview streetcar south through the Lever property to link to the Portlands LRT.
The city planners mentioned residential developments north of Queen in the core area as a reason to keep the route as far north as possible. And several on this site want to head south along Wellington or King, compromise and use Richmond or Adelaide. Preserve the King and Queen streetcars intact during the construction. A stop at NPS would be handy (Grab your torch, pitchfork and TTC token – add politician as desired!) but the potential for redevelopment just a bit further south by the private sector is greater. Staying north of King gets separation from Union, spreads the load out in the PATH system and gives GO or ST riders a reason to switch to RL at Gerrard rather than Union to make a closer final transfer. Keeping the line to only 7 [stations] (Pape, Gerrard Sq., Queen/Broadview {Munro St.}, Opera Center (Osgoode/Bay), Yonge/Victoria, Church, Moss Park/Stonecutter) or 8 (St. Paul’s) stops would keep this line shorter than the BD/Y Pape to Queen alternate.
Steve: I must declare a conflict-of-interest if, per the new Metrolinx station naming protocol’s prohibition on duplicate station names, a Queen/Broadview station is proposed called “Munro” after the street just to the west. I prefer “Dangerous Dan’s Station”, and was never happy that “York Mills” lost its originally planned named of “Hogg’s Hollow”.
One interesting post-it note suggestion was to extend the streecar tracks north up Bay from Union Loop to the new DRL station on Bay. Why stop there? Run further north, past City Hall and join up with the old tracks on Bay, allowing connection to the Bus station, the Dundas and Carlton cars. With a station on Church, a downtown core circulator loop could work.
LikeLike
Steve said that the Broadview options don’t fare well and that Pape was the only real choice but gave no reason for it. Does anyone know why or is it just his personal preference like mine is Broadview as I live on Broadview? Broadview is also much better than Pape in the sense that the father you go from Downtown, the lower the density and ridership.
Steve: If you read the background paper, you would know that I am reporting on what it says. The Broadview options do not fare as well as the Pape options in the scoring, and they are dropped from further consideration. By the way, I live at Broadview and Danforth, and so a Broadview alignment would, in theory, be “better” for me if one counts “better” as having my neighbourhood torn apart for a major subway node. Alignments via streets further east, including the Donlands alignment I have written about in the past, have already been dropped in earlier rounds of the study.
The specific points on which Broadview scores differently from Pape include:
A problem common to the evaluation of both Broadview and Pape alignments is that some issues (pro and con) are cited that do not apply to all variants for each corridor. For example, a Broadview/Dundas route serves Regent Park better, but a Broadview/Queen route provides the same access as Pape/Queen. Also, serving the Unilever site is better for the Pape options that are routed further south, but these are rejected because of the more difficult river crossing. There also appears to be a presumption that there would be no Broadview streetcar southerly extension (indeed no streetcar on Broadview itself), but this is not stated explicitly.
LikeLike
Will the TTC still need 60 new streetcars now that the DRL is a sure thing as confirmed by it finally being mentioned in the provincial budget? I think that the eastern portions of many parallel east-west streetcar routes would be absorbed by the DRL as with the Bloor-Danforth subway 50 years ago. The western streetcar roues should survive until DRL west is built but that is unlikely to be built before 2025.
Steve: First off, the (D)RL is no sure thing even to the east and it does not replace demand in the heavy Broadview/King corridor. As for an RL West, we won’t even see the eastern leg under construction until 2025 at best case, and a western leg is much further off. I have little doubt those cars will be needed.
LikeLike
Steve, what are your thoughts on the seeming omission of a station at Parliament St? Isn’t Jarvis/Parliament/River more practical than just Sherbourne/Sumach?
Steve: By analogy to Bloor Street, there are only two stops between Broadview and Yonge. One on the west side of the river and one at Sherbourne. That said, Sumach is an odd choice, but it ties in with the idea that a more northerly subway (Queen) serves Regent Park. The rationale for both the alignment choice and the station locations is forced.
LikeLike
Could someone explain this one to me? Eastern Avenue is straight once you get east of the bend near Broadview. There’s plenty of room for a 150m station box.
Steve: This assumes that the RL follows the rail corridor, not Eastern Avenue. But yes there is a lot of confusion about which constraints and benefits apply to which version of the alignment. I cannot help having the feeling that the fix was in for the Pape to Queen via Queen alignment, and the arguments were made to favour that choice. Unfortunately, project staff are still quoting remarks that only apply to certain sub-options as if they were generic, which they are not.
LikeLike
It was also odd that the only time there was mention of “a lot of expropriation” was about the route under Eastern Ave. Most of the land on the south side of Eastern, between Booth and Lewis belongs to the city! There’s an Enbridge maintenance yard and the building currently housing the Avenue Road furniture store, but the two beautiful, big, old buildings on the south side are both city owned. There are currently fewer building on the south side between Booth and Pape than there are on Queen. So the notion that expropriation is more of a problem here than elsewhere is a head-scratcher. I think Steve’s right and that the fix is in for a Queen St. alignment, unless minds can be changed. The engineers might prefer that route, but the planning department has to consider city building, and most of the expected growth in the area is between Queen and Lakeshore, which makes the Eastern Ave route all the more compelling, IMHO.
Steve: I think the reference to expropriation is for the curve taking the line out of Pape into Queen or Eastern, and the different types of building in the two locations.
LikeLike
Dear Steve, did you read the reports carefully? Who told you that the DRL might be pushed south to Wellington? Only Queen and King are under consideration – I personally would prefer King as I don’t want Queen St changed beyond recognition by a subway. The DRL will change beyond recognition every neighbourhood it goes through and so I would much rather it go through King than Queen.
Steve: There has been mention of both King and Wellington as alternatives for the “King” alignment just as there has been of Queen and Richmond. Have a look at the Pape-King alignment info where Wellington is mentioned as an alternative. I appear to have been reading the reports more carefully than you.
LikeLike
Why exactly can’t you build a subway station on a curve in this case? Especially since the radii of curvature in the line sketch appear very large compared with the train length. Union Square Lexington Line platforms in NYC come to mind.
Steve: The problem is that the platform cannot be placed close to the train due to the curve. Gap fillers are required. They are possible, but should be avoided.
LikeLike
I worry that we are not looking hard enough at the obvious questions going forward in terms of planning. The most southerly alignment does a better job of supporting a future Western commuter station, and the Waterfront development, however, in not discussing the logical continuation, we are not looking at how we will handle additional growth from the west. Do we think now in terms of a real link for LRT from the west? Do we look at support for the CNE grounds, and substantial development there? It seems to me there is too much attention on Queen, although, making the end of the Queensway is an interesting goal, points between need to be equally considered, and the route for a DRL from the East needs to account for where it should have future extension in the west.
LikeLike
The alignment of the DRL through downtown must be chosen to come as close to the final destination as possible. Nobody will transfer to DRL at Pape and then transfer again at Yonge. Choosing Queen will likely mean that anyone wanting to go to Dundas or King will view the DRL as too inconvenient and chose the single transfer at Y-B instead. An alignment between Queen and King (i.e. Richmond) would be perceived as serving both Queen and King – and likely would be much more popular.
LikeLike
First a question that occurred to me – Which ‘soft site’ are they referring to at the SW corner of Queen and Broadview? The properties currently being developed by Streetcar Development? (Healthy neighbourhoods criteria 2)
Steve: That corner is a convenience store and a few restaurants. They won’t be missed. The Streetcar Development properties are further west.
I see valuable points on both sides generally on the Queen vs. King alignment debate, but leaving that aside to look at the two Queen-Pape options I cannot understand the choice of the under-Queen route.
To me it has to be assumed that significant and almost continuous disruptions of the 501 for a period of 5-10 years (depending on staging – but take a look at the forest of rigs on Eglinton over the last 3 years to get an idea of what I mean) should be taken off the table as unacceptable, so the Richmond – Eastern alignment is preferable and would serve the Unilever lands.
I question the value of a Queen/Pape station. The area is already well served by streetcar and may not actually welcome the increases in density and new built form that will accompany a subway. At north of $150 million per station it seems that skipping a stop there in favour of greater speed and less disruption would be better. If the community needs improved transit, a fraction of that invested in improvements to the streetcar and other public realm would likely be better.
If we remove the Queen/Pape stop, then the most direct path may be to follow the railway right-of-way to Gerrard Square and then north up Pape.
Fewer stops will increase speed and the streetcar network is great for local distributor service.
Thoughts?
Steve: I agree with your analysis. Indeed a Pape/Queen station is almost impossible on a Pape/Queen alignment because the curve will take the subway diagonally away from the intersection and the station, if any, would actually have to be at Carlaw. Bluntly, I think the evaluation has been cooked to keep the RL out of the way of SmartTrack.
LikeLike
There’s so much grade A bull in the evaluation. In particular in the Experience category, where Queen received full marks (4/4) for access to what the study calls key destinations while King received half marks (2/4).
For example AGO is over 1km away from a City Hall station but is counted as a plus. Supposed access to hospitals which are well north of Dundas and even at College get credited in favour of Queen. Why? Damned if I’d know!
Destinations (such as the aforementioned hospitals) Eaton Centre, and Ryerson which are already well served by multiple subway stations score points for Queen and seem to be weighted the same against locations on King which are not connected to the rapid transit network at all.
Using some of their bullshit criteria and evaluation rationale I struggle to see how a station at King and Bay doesn’t count as access to City Hall, Eaton Centre, Ryerson or places like Southcore or how a station near the Distillery doesn’t count as a plus for Regent Park.
I haven’t been at any of the meetings, I missed the one I was planning on going to due to work commitments, but I hope people have been tearing into their fradulent analysis.
LikeLike
Another example, a station at Front East and Cherry is actually closer to the Daniels Spectrum on Dundas than a City Hall station would be to some of the hospitals listed as key destinations.
The city evaluation is seriously suspect.
LikeLike