Toronto’s Public Works and Infrastructure Committee (PWIC) will consider an updated report on the Gardiner East reconstruction options at a special meeting on May 13, 2015 where this will be the only item on the agenda. (Note that additional detailed reports are linked at the bottom the main report.)
There has been much discussion of the alternative designs for the expressway section between Jarvis Street and the Don River and, broadly speaking, there are two factions in the debate.
- For one, the primary issues are to maintain speed and capacity of the road system, and to avoid gridlock.
- For the other, the primary issue is the redevelopment of the waterfront, and the release of lands from the shadow of the expressway structure.
Both camps seek to encourage economic growth in Toronto, but by different means and with different underlying assumptions.
A further issue, largely absent from the Gardiner debate, is the role and comparative benefits of various transit projects ranging from GO/RER/SmartTrack at the regional level, down to subway options including the Scarborough Subway Extension and the Downtown Relief Line, and local transit including the Waterfront East LRT line and a proposed Broadview Extension south across Lake Shore to Commissioners Street including a Broadview streetcar.
In early 2014, four proposals for the Jarvis-to-Don segment were before Council:
- Maintain the existing expressway
- Improve the expressway with a new structure reduced in width (to acknowledge the absence of the Scarborough Expressway as a feeder)
- Replace the expressway with a completely new structure whose footprint at grade would be less than the present design
- Remove the expressway and create a consolidated Lake Shore Boulevard similar to University Avenue
Both the Improve and Replace options would involve reconfiguration of Lake Shore Boulevard under the expressway to the extent that the expressway structure would allow.
When this was before Council, there was much concern about the travel time estimates and congestion in general. Some of the projected times did not, on their face, make sense to Councillors or the change in travel times were unacceptably large.
Further issues arose from an alternative “Hybrid” proposal put forward by First Gulf, developers of the former Unilever site east of the Don River. Access to their property is constrained by existing structures, notably the residual “Scarborough Expressway” branch which descends to grade at Logan Avenue and Lake Shore Boulevard east of their site.
The Hybrid scheme eliminates the Scarborough ramps so that the merge of expressway traffic into Lake Shore occurs west of the Don River rather than east of it. This opens up the south side of the First Gulf site.
As for the Gardiner/Don Valley Parkway ramps, these would be shifted so that the Gardiner, instead of swinging south along the edge of Keating Channel, would lie further north adjacent to the existing GO Transit Don Yard.
Council sent the whole package back to the planners asking that (a) they review travel time projections and (b) report on the Hybrid Option. It is this report that has now come back to Council.
The 2015 Proposals
There are now only three proposals on the table:
- Maintain the existing expressway (this is the “base case”, or the “do nothing” option)
- A modified version of the Hybrid option
- Remove the expressway (this variation is now called the “Boulevard” option)
As proposed, the Hybrid option was unworkable for two major reasons:
- The transition between the Gardiner and DVP would have required a very sharp curve that could not be driven at expressway speeds. This would form a bottleneck to road traffic and would be a safety hazard because of the sudden transitions between high speed expressway driving and a much slower curve.
- The proposed alignment conflicts with a Toronto Water facility east of Cherry Street and a sediment control basin for the Don River.
The Hybrid has been reworked so that it follows the same curve as the existing Gardiner/DVP link, but the Scarborough branch splits off with new ramps west of the Don merging into Lake Shore Boulevard.
For those who only know the area as it is today, or are unfamiliar with evolving plans for the eastern waterfront, there are several points to note in the illustrations above:
- The ramps down from the Gardiner to Lake Shore begin west of Cherry rather than east of the Don River as they do today. Lake Shore Boulevard would cross the Don on a widened bridge, and east of the river, there would be no elevated structures.
- The ramps between the Gardiner and the DVP are rebuilt in their current location starting from the elevated structure immediately north of the Keating Channel and swinging across the river to “land” on the DVP just south of the railway bridge that carries GO Transit operations east from downtown. This bridge cannot be moved because doing so would require complete realignment of the rail corridor. The railway berm doubles as part of the flood protection system.
- Much of the development shown above does not yet exist, and the form of the Don River is completely changed to its new “naturalized” state. South of the Lake Shore bridge, this takes the river south parallel to the Don Roadway and then west out into the lake. Northwest of the DVP ramps is a sediment control area for the river. Some of the buildings west of Corktown Common (the park north of the rail corridor and west of the Don) have been built as part of the Pan Am Games Athletes’ Village, but more are to come. The land south of the rail corridor is only just beginning to develop, and that activity is at the western edge of the illustration. The area south of the Keating Channel, known as “Villiers Island” is now mainly vacant. The road at the south side of the development shown above is Commissioners Street.
- Two new bridges will be added linking across the Keating Channel. One is at Munitions Street and the other is at a relocated Cherry Street that shifts west from its current location. The new bridge will not be a lift bridge because there will be no need to provide clearance for tall vessels into this area.
- Transit connections via a realigned Queens Quay and New Cherry Street will serve the Port Lands including Villiers Island. Not shown on the map is a possible connection east via Commissioners to an extended Broadview Avenue.
- From a neighbourhood design perspective, the alignment of the Gardiner along the north side of Keating Channel limits the options for opening up this area as a continuation of the planned Waterfront Promenade that is already an integral part of developments further west.
The Boulevard option has undergone minor changes to improve intersection operations between Cherry and Jarvis Streets in an attempt to reduce the travel time penalties of this configuration.
Important notes about the Boulevard option are:
- In the Boulevard option, all of the traffic flow east of Cherry stays further north than with the Hybrid option on a consolidated, 8-lane Lake Shore Boulevard. This splits at the DVP giving two lanes in each direction to the DVP and to Lake Shore east of the river matching the current configuration, but with much less elevated structure, and with the road at the water’s edge only for the Don River crossing.
- If the DVP ramps stay in their present location, land west of the Don and north of Keating Channel that was expected to be available for development on the water’s edge will be occupied or hemmed in by them.
- The Boulevard option, combined with other plans for the West Don Lands, East Bay Front and Villiers Island, establishes the entire area as one of “local” streets rather than as one dominated by the expressway structure.
The major tradeoffs come in which vision of the waterfront, and of downtown Toronto in general, is the priority for citizens and politicians.
Travel Time Estimates
The big debate over which option to choose will be a battle of competing demand and travel time models. One view is in the reports from the city staff and consultants, while another is in a study commissioned by the “Gardiner Coalition” from the Transportation Research Institute at the University of Toronto. The Coalition includes the CAA, Courier, Motor Coach and Trucking Associations, Redpath Sugar and the Financial District Business Improvement Area (BIA). [Note: The link here points to the cp24.com website where the report is available as part of their coverage of this issue. It does not appear to be online elsewhere, and the Coalition does not appear to have its own web presence. The UofT site announces that the study has been done, but does not provide a link to any documents.]
The City Studies
City studies have provided two different sets of estimates. The first was in the February 2014 report that triggered the current round of debate.
Note that travel times have already grown from locations to the northeast that are served by the DVP between 2001 and 2012. This is due to added congestion on the DVP that is not affected by the Gardiner options which have not yet been implemented. The DVP’s capacity is constrained further north with a considerable amount of traffic leaving at Eglinton and at Bloor. By 2031, travel times are expected to rise by another five minutes even with the current structure for trips originating on the DVP or Gardiner West.
During the 2014 discussions, attempts to understand the numbers in the chart above proved quite frustrating. For example, the Improve and Replace options produce longer travel times from all origins than the Maintain option, but the reason for this was never explained.
The compound effect of all of the increases between 2012 and 2031 Remove is quite substantial, and options that retain the Gardiner do only slightly better. The fundamental issue is that there is only so much road capacity to go around, and as long as there is growth in road travel, congestion will get much worse even if the Gardiner stays in an elevated configuration.
Another important issue is the growth in travel times from location “C” in the map, the southwest corner of Scarborough. Much of this traffic flows along Kingston Road, Woodbine and Lake Shore merging into the downtown street system at the Don River. For many drivers, location C is a midpoint, not the start of their journeys, and their total commute times from other parts of southern Scarborough will be affected by road capacity limitations there too. Travel times from “A”, “B” and “E” on the map are affected by existing capacity constraints moreso than from “C”, but that eastern approach via Lake Shore is filling up. This is demand the Scarborough Expressway was designed to address, but like its cousin, the Spadina Expressway, it would have done more harm than good.
Eventually, motorists have to accept that growth cannot be achieved by auto travel. That’s fine provided that there is an alternative.
In 2015, the projected travel times have changed somewhat.
The “2031 Base Case” above is the “Maintain” option of 2014, and right away we can see that the projected times have changed slightly. This should be explained, although it may simply be a case of rounding in the 2014 report where all values are multiples of 5.
For DVP and Gardiner West origins, the Hybrid option has the same travel times as the Base Case, unsurprising given that it is only a slightly modified version of what is there today. The Improve and Replace options are no longer on the table, but the question remains from 2014 of why their projected times were higher than the Maintain option, and more importantly why the 2015 Hybrid option performs better than those two abandoned schemes.
Where the Remove (Boulevard) option was 10 minutes higher than the Maintain option in 2014, this difference has been wrestled down to only 3 minutes through design changes. This is rather difficult to believe given the relatively small portion of the Gardiner East that has been “tweaked” in 2015, and the study authors would have done well to explain this in greater detail. Their failure to do so, as in the 2014 report’s lack of detail, undermines the credibility of the Boulevard option.
Again the “C” origins from southwest Scarborough face longer travel times regardless of which option is chosen.
The Gardiner Coalition Study
This study examines two variants of the Remove option, but does not consider the Hybrid scheme. The principal difference between the variants lies in the treatment of competing demand for road capacity. In one variant, the focus is on pedestrian movement, while on the other the focus is on auto traffic. This affects design elements notably signal timings where the two types of traffic will conflict.
Simulations of traffic flow (pp. 15-18) for each configuration show that there is little change in traffic volume between the Maintain and Replace options. This is related to another observation in the City’s study which found that demand on the expressways is not affected by the presence or absence of transit such as GO improvements, SmartTrack, the DRL or the Waterfront LRT line.
These findings are not surprising given that the traffic flowing into downtown from the DVP is a very small fraction of the total peak demand, and the road network has been “full” for years. With new transit, some existing trips may divert from the expressways, but these will easily be backfilled from existing demand. Without new transit, there is no capacity for the road network to absorb more trips, and so transit gives the impression of being “without effect”.
What is missing both in the Gardiner Coalition and City studies is any discussion of the scale of growth that is expected on the transit system and various proposed components of that network. I will return to this point later.
Both the City and Coalition studies use projected traffic volumes on the DVP, Gardiner and Lake Shore as their starting point. There are differences between the numbers, but generally speaking the volumes are comparable. In both models, approximately 7,400 vehicles arrive via the DVP and Lake Shore, and they are dispersed to various destinations. The demand on the segment between the DVP and Jarvis Street is almost identical in both models.
City map of demand distribution westbound:
The Coalition study has similar figures for these flows:
The Coalition study also includes a chart showing traffic that joins the Gardiner westbound at Jarvis and at Bay/Spadina. These add considerably to traffic west of Jarvis, but that section of the road is not under discussion. However, this volume must be factored into the behaviour of intersections, notably at Jarvis and the Gardiner/Lake Shore.
For eastbound traffic, the basic numbers used by both the City and Coalition studies are similar, but there is an important difference in that the Coalition also shows eastbound traffic joining the Gardiner at downtown access points, not just the traffic coming east from Parkdale and points west.
City map of demand distribution eastbound:
Coalition map of eastbound demand without traffic originating downtown:
Coalition map of eastbound demand including downtown traffic:
Unfortunately, the Coalition study shows these added flows, but does not discuss how they affect the modelled behaviour of the road network notably the degree to which the Remove option might cause traffic to back up into the higher speed sections of the DVP and Gardiner approching the revised central section.
Among the design options included in the Coalition versions of the Remove option are:
- Addition of the Richmond/Adelaide bike lanes
- Different speeds and intersection signal timings to favour traffic, or to favour pedestrian activity, notably north-south crossings of the Boulevard.
There is no explanation of why the bike lanes were not included in the Base Case model if they are to be part of the future road network regardless of what happens to the Gardiner. Taking some capacity away from Richmond westbound could produce queueing back onto the southbound DVP, but this is not by itself an effect of the Remove option. These two factors need to be disentangled. An obvious related question is whether the City models included the bike lanes, and if so what was their effect on the behaviour of each option.
The simulated flows for the Gardiner eastbound in the Coalition study are intriguing because they show a reduced flow for the Remove option with pedestrian emphasis whose effect stretches west into Parkdale. This implies that traffic will be unable to exit as efficiently as it does today at Jarvis and points west to downtown because of backlogs of traffic pushing west from Jarvis.
Another way of looking at the demand modelling is to review trip times for various trips through the core area. The Coalition report shows this in the following diagrams.
For westbound trips, the difference between the base case (Maintain) and both of the Remove scenarios is roughly the same for all destinations: 3-4 minutes for option 1 (traffic oriented signalling) versus 5.5-6.5 minutes for option 2 (pedestrian oriented signalling). This implies that all of the added delay occurs east of Sherbourne because the delays do not grow as one moves further west. However, it does not explain where the extra time is actually spent including the effect of reduced capacity on Richmond due to the bike lanes in the Coalition’s model. If these lanes were added to the Maintain scenario, what would happen? Would they create a delay on the DVP independently of whatever happens on the Gardiner?
For eastbound trips, the spread between the Base Case and the Remove options grows as one moves further east, and for Remove Option 2 there is an added 2.5 minutes even for trips bound for the Spadina off ramp. This implies congestion reaching back across the core that adds more time to each trip the longer one spends in that segment. The extra trips that join the eastbound flow at downtown ramps have to be accommodated by the redesigned eastern section of the Gardiner, and the City model does not take them into account.
Updated May 12, 2015: In a comment, Karl Junkin has noted that the Remove Option 1 in the chart above shows lower times for the “high” and “average” values than the Base Case for traffic bound to the York, Yonge and Jarvis off ramps (albeit a small change).
These effects require further study and explanation.
It is almost counter-intuitive that a change on the eastern portion of the Gardiner would have effects on flow in from the west, and it will be important to learn if the geometry of the modelled Remove option matches what the City’s Boulevard option actually proposes.
The Coalition calculates a value for the time lost through added trip times through the core on a reconfigured Gardiner, and their clear aim is to show that this is a “cost” that must be weighed against capital (and future maintenance) cost of any Remove scenario. The challenge in such evaluations is that the “savings” of a faster road accrues privately (time and fuel savings) while the cost is a public one funded by taxes.
What is missing from the discussion is the question of competing demands on capital and operating budgets for transportation facilities that would serve far more trips than the comparatively small number who enter downtown via the Gardiner.
At great expense, we could make commuting times better (or at least not much worse than they are today) for Gardiner users, but what about the 68% of trips that are served by transit (2011), a figure that will grow over the decades because almost all new trips to the core will be on transit. There simply is no place left to put them on roads. What will be spent to serve this growth?
Competing Demands For Capital Investment
- Gardiner East Project
- GO/RER: Upgrade of the GO network to support at least a 15-minute all day service on most or all branches, with more frequent service on some corridors.
- SmartTrack: A proposed frequent service also on a 15-minute headway proposed to run from Unionville to the commercial district south of Pearson Airport via the trackage used by the Stouffville and Kitchener services, branching off at Eglinton into a separate tunnel westward. This scheme will doubtless run into implementation and funding problems because of track time constraints on the shared GO corridors, and the dubious (and expensive) western leg via Eglinton. This could very well wind up as a subset of GO Transit operations, possibly with its own logo and fare structure depending on how much Mayor Tory has his heart set on implementing a unique service that sprang from his campaign.
- DRL: A subway line with three possible segments: (1) a core link from the Danforth Subway to the vicinity of Union Station, (2) a north-easterly extension to Don Mills & Eglinton, and (3) a northwesterly extension linking to the Bloor Subway. This is an expensive project whose purpose is to divert demand from the main subway interchanges at St. George and Bloor-Yonge. This provides added capacity to the north-south lines and reduces the need for complex changes to increase the capacity at those interchanges.
- SSE: A subway line running from Kennedy Station to either Sheppard Avenue or to Scarborough Town Centre via a route that is still to be decided. This line directly competes with the Stouffville GO/SmartTrack corridor.
- Waterfront East LRT: An LRT line mirroring the existing Harbourfront route on Queens Quay West that would eventually operate south into Villiers Island and potentially hook up with a Broadview extension and with the already-built spur on Cherry Street that ends at the rail corridor.
- Broadview LRT link: If Broadview Avenue is extended south to Commissioners through the First Gulf site, a direct service from the Danforth Subway at Broadview Station to the eastern waterfront would be possible.
There will almost certainly be a GO station to serve new development at the Don River, but GO/RER (and SmartTrack) cannot provide the fine grained service to address new employment and residential demand in the eastern waterfront. Their role is to provide for metropolitan and regional travel into the core, the same type of demand that the expressway network serves. At a finer grained level, both the DRL and any new waterfront transit would handle the local trips including links to the regional network.
In a few decades, the eastern waterfront may be home to 50,000 in an area that has seen little development since an industrial heyday beyond the memory of most Torontonians. Tens of thousands more will work and go to school there. Where is the debate about how all of these trips will be served, how the neighbourhoods will function without good transit? An elevated Gardiner is almost a non-entity for local travel to, from and within a small new city on the edge of downtown.
All of the “Gardiner” discussion has concentrated on the time that would be lost for road traffic, but the wider question for any investment is where it should rank against other possible uses of limited capital. Studies are now underway to determine the relative performance of options for most items listed here, but these will not be available until Fall 2015. In a rush to “fix the Gardiner” and to placate road users whose primary concerns are congestion and gridlock, we might launch down a path that precludes more investment in transit and a recognition of its vital role in the economy of downtown Toronto.
One must also ask why transit projects must beg and compete for funding when keeping the roads flowing seems to be essential at any price.
What Will It Cost?
Media coverage of the costs might leave some readers wondering just how the numbers are counted. In brief, the following considerations apply:
- The cost to build each of the options ranges from $396-million for the Maintain scheme up to $524m for the Hybrid with the Remove option slightly higher than Maintain.
- All three options will cost substantially more than the $232m now set aside in future capital plans by the City, and this will challenge the City’s overall debt management strategy.
- Estimates for the Hybrid and Remove options have a ±20% margin because these have not been subject to detailed design, whereas more work has been done on the Maintain option and its estimated cost is felt to be accurate within a ±10% margin.
The debt management issue is a particularly troublesome one because current projections show Toronto hitting its self-imposed ratio of debt service costs to tax revenue of 15% by 2020 leaving no headroom for added borrowing for a multitude of projects. Either a very generous Tooth Fairy will appear to resolve this problem, or the City will have to make difficult choices about raising taxes to support additional capital debt.
Construction is only part of the story, however, because elevated expressways cost more money to maintain than roads at grade.
The future cost of alternatives using an elevated structure are substantially higher than the Boulevard option because, as the Gardiner regularly reminds us, elevated roadways do not last forever. Even with improved road building technology, there will still be higher costs.
This chart has appeared quite regularly and the $919m Hybrid price is often quoted in comparison to the $461m for the Boulevard. However, much of the maintenance is in the future, and for present day comparisons, this is reduced to “Net Present Value” – the amount of money you would have to invest today so that it would pay all of the future expenses. Even the construction cost is lower on this basis because the work will not actually start until 2020.
There is still a difference between the options, but the numbers are not quite as imposing.
Any comparison to transit projects must also look at costs, benefits and when these will occur to put the proposals on an equal footing. For example, the Gardiner Coalition prices out the cost of time for vehicles delayed in traffic, but we have heard nothing about the value of time for the much higher number of riders (present and future) who will attempt to use the TTC and GO systems.
The Rush To A Decision
The City’s study will come to the Public Works and Infrastructure Committee on May 13, and then to Council on June 10-11. In some political circles, the imperative is to “decide now” because (a) the expressway will fall down at any moment and (b) the wind appears to be in the sails, so to speak, of those who want to keep an elevated roadway in whatever form.
However, there are many unknowns, notably the assumptions, validity and implications of competing traffic modelling exercises, and the need to understand how any expressway configuration will behave as part of a larger downtown road network.
Probably the most important unknown is the future role of transit: where will it be, what types of demand will it serve, will getting to the core be easier for someone in Unionville or Bramalea than for someone in Scarborough, The Beach, West Toronto or Etobicoke? What will happen if we don’t expand the transit network to meet demand?
Will Toronto rise to the challenge of providing service to its growing, dense downtown business district, or will a pre-occupation with a relatively small commuting demand hi-jack the very necessary debate on the limits of funding, not to mention the dwindling role of road traffic? Will the future prosperity and attractiveness of downtown be defined by a chunk of expressway that serves a small minority of commuting traffic, or by a commitment to very substantial additions to the transit network’s capacity to deliver riders to those shining towers so favoured by those who are Toronto boosters?
Council should not be rushed to decide immediately on a Gardiner option until the studies now underway for transit options are on the table. This is a short delay in the time scale of all of the projects, but it gives us the opportunity to look at them all, to decide where those precious taxpayer dollars should go.
This article has deliberately avoided two components of the debate and reports: the questions of city building and urban design as these would be influenced by whichever Gardiner survives into the next decade, and the Goods Movement Study that addresses the needs of commercial traffic both from the Port Lands and other industrial areas east of downtown, and for access by commercial traffic to the core in general.
I wanted to concentrate on the transit vs road tradeoffs, and what, to me, is an undue focus on one component of a very complex network. No matter what we do, making everyone happy is simply not possible. Today’s squeaky wheels represent too small a part of the total population and travel demand. They should not dictate the outcome of the larger discussion.