Updated on January 28, 2017:

Changes include:

• Addition of data for November and December 2016.
• Reformattied chart pages so that data for years 2014, 2015 and 2016 appear on separate sheets.

With the launch of Toronto’s TOCore project, the city set in motion a complete rethink of what “Downtown” means and how it will evolve in coming decades. On some counts, one might argue that this work is long overdue as concentration of office and residential space in a very small area brings many problems for residents and businesses, not to mention a very competitive demand for a crucial resource – road space.

I will leave the debate on many of these issues to other people and forums, but as this is a transit blog, my focus is on understanding how transit works (or doesn’t) and what effects might result from various proposals.

In the Globe & Mail, Oliver Moore writes about “Complete Streets” and how this design philosophy could affect Toronto. Without question, better attention must be paid to improving the safety and usability of major streets by pedestrians (who are also transit riders) and cyclists who collectively outnumber the motor traffic.

King Street has long been the busiest of the downtown streetcar routes carrying about 65,000 riders every weekday. But these riders do not all travel to and from the business district at King & Bay, nor do they all travel in conventional am and pm peak commuting times. New demands on the shoulders of downtown such as Liberty Village and the St. Lawrence district include not only residents bound for jobs at Bay Street, but workers and students headed to offices and schools on counter-peak trips. Indeed, the term “counter peak” can seem odd when one looks at some of the demand patterns.

In 2014-15, I was retained by the City’s Transportation Department and the TTC to review the major streetcar routes with a view to identifying locations on the shoulders of peak periods where parking and turning restrictions should be extended beyond the traditional two hour window. As a result of this and other surveys conducted by the City, traffic regulations were changed in several areas. This brought some improvement, typically eliminating anomalies where the pm peak, for example, actually was worst for transit service in the hours just before and just after the “official” rush hour.

However, that review was considerably smaller than the goals of TOCore. A redesign of a street like King is an all-day effort, and one that could, depending on its scope, affect a great deal of the streetcar route. This is not a case of tweaking a few hours a day, but of reinventing the street.

Although studies for the City began with data from fall 2013 and early 2014, I have been tracking the operation of several routes from time to time using data provided by the TTC over many years, for which much thanks.

The material presented here is a continuation of my earlier work, but expanded both in space and time. For a detailed description of the methodology used here, please refer to Methodology For Analysis of TTC’s Vehicle Tracking Data.

504 King is a special case being the subject of City and TTC studies, and my data archives are more substantial for this than for other routes. For 504 King this includes:

• December 2006
• November 2011
• March 2012
• May and July-December 2013
• All months for 2014 and 2015
• January and March-December 2016

If there are improvements to be had on King, an obvious starting point is to ask whether, where and how travel times for transit vehicles have changed over the years. How does the route operate under “best case” situations when competing traffic is low and streetcars are not overcrowded causing loading delays?

The King Route as a Whole (Updated)

Route 504 King operates between Dundas West and Broadview Stations via Dundas, Roncesvalles, King, Queen and Broadview, and has followed this route since July 1, 1923 when the termini were Vincent and Erindale Loops near the present-day subway stations. Demand on the route is broken into many overlapping sections including the north-south links on Broadview and Roncesvalles, as well as traffic to western Parkdale, Liberty Village, St. Lawrence and more recently the Distillery District.

First I will look first at the central section between Parliament and The Queensway/Roncesvalles, and then turn to individual portions of the route.

• 504 King Parliament to Queensway WB Summary
• 504 King Parliament to Queensway WB By Year
• 504 King Queensway to Parliament EB Summary
• 504 King Queensway to Parliament EB By Year

The summary charts contain data for all years so that the overall patterns and evolution is visible, subdivided into 6 pages covering three-hour time periods from 6:00 am to midnight.

The yearly charts separate the data into 24 pages in total for individual years (2013 and before, 2014, 2015 and 2016) and the six time periods.

• Within each page, average travel times are plotted in a different colour for each half hour. The time represents the point where a vehicle entered the section (e.g. Parliament westbound, or King east of Roncesvalles/Queensway eastbound). Therefore, data for “6:00” represent trips beginning within the period 6:00 to 6:29.
• Where there is a break in continuity of dates, there is also a break in the data lines.
• The solid lines show the average values, and the dotted lines show the standard deviations (SD) of these values. The lower the SD value, the more consistent the individual values making up the averages are. Spikes in the SD values indicate that for the half hour and week where this occurs, the average values were not clustered probably due to a delay, or because traffic conditions made travel times less consistent.

Major events affecting the route included:

• August 2013: Reconstruction of King/Spadina intersection. All service diverted via Church, Queen and Shaw. The westbound diversion was changed to use York Street part way through this period.
• September 2013: Immediately following TIFF, construction began on Adelaide Street causing eastbound traffic to shift south to King.
• November-December 2013: Hydro work on King occupied the eastbound curb lane at various places near University Avenue.
• April 2014: Reconstruction of the Gardiner Expressway commenced in late April and continued through the year with spillover traffic effects on King.
• September 2014: Toronto International Film Festival (TIFF) mall operated with streetcars running through via King in semi-protected lanes.
• September 2015: TIFF diversion via York (WB), Church (EB), Queen and Spadina.
• Late March to early April, 2016: Reconstruction of King/Charlotte intersection. Same diversions as were used for King/Spadina in 2013.
• May 8 to December 9, 2016: All 501 Queen service diverted via Spadina, King and Shaw adding delays to 504 service at turning locations.
• September 2016: Route split for opening days of TIFF. Service operated through TIFF district for Week 2.

Travel times at the end of the year generally drop due to lighter traffic in Christmas Week. Note that when special days such as Christmas, Boxing Day and New Year’s fall on weekdays, these are not included in the averages. (Statutory holidays are similarly excluded during other parts of the year.)

Averages for a long route segment can mask local effects along the way as operators will tend to make up for delays in one area with faster operation in others. However, a striking point about all of these charts is the degree to which travel times overall have not changed much in 2016 relative to recent years, although there has been some growth notably in the afternoons and early evenings. “Older” data must be viewed with care taking into account the period it covers. For example, the 2006 data are from December and include the usual year-end lull. Other periods include effects from long-running construction projects.

Some portions of the route show little change in travel times over the years, while others show major upheavals. In areas where population and traffic are increasing, the question is “how much capacity remains”. Sorting all of this out in the context of a new street design can be challenging, but an important aspect is that the street will rarely be operating under “ideal” conditions.

Route Segments

In the following sections, data for each segment of the route are presented separately to show how changes in travel times are localized. A few notes:

• The vertical scale on these charts is not always the same in order to accommodate spikes in values while preserving the general “feel” of the data.
• Values by half hour are based on vehicles entering the segment, not the full route, during each period.
• The charts are consolidated onto a single page rather than over two as above, but the format is otherwise identical.

Parliament to Jarvis (Updated)

• 504_Parliament to Jarvis WB Summary
• 504_Parliament to Jarvis WB By Year
• 504_Jarvis to Parliament EB Summary
• 504_Jarvis to Parliament EB By Year

Travel times between Parliament and Jarvis vary through the day as one would expect – higher during peak periods, lower otherwise – but in 2016 they are similar to values years ago.

Jarvis to Yonge (Updated)

• 504_Jarvis to Yonge WB Summary
• 504_Jarvis to Yonge WB By Year
• 504_Yonge to Jarvis EB Summary
• 504_Yonge to Jarvis EB By Year

Travel times are unchanged over the years with the notable exception of a spike in August 2013 when the service diverted via Queen. The westbound diversion operated via Church, Richmond, Victoria and Queen initially, later via King to York and then north to Queen. The eastbound diversion operated via Queen and Church throughout the construction at King & Spadina.

The travel times are measured from Queen & Yonge to King & Jarvis where cars were on diversion. Because the route reverted to King Street over this segment westbound before eastbound, the spike in travel times is not as wide for that direction.

No change in average westbound travel times is evident from the removal of the stop at Victoria Street late in June 2016.

Yonge to University (Updated)

The segment west from Yonge Street through the financial district shows more change between periods of the day, and greater effects when service is disrupted. The most severe problems arose in late 2013 eastbound due to curb lane work by Toronto Hydro. This tripled the travel times during the height of the pm peak.

The diversion via Queen for Spadina/King construction affected travel times during August 2013, as did the diversion for Charlotte/King in March-April 2016.

No change in average travel times is evident from the removal of stops bothways at York Street late in June 2016.

• 504_Yonge to University WB Summary
• 504_Yonge to University WB By Year
• 504_University to Yonge EB Summary
• 504_University to Yonge EB By Year

University to John (Updated)

This section of the route has been affected by diversions for track construction as well as general slowdowns caused by utility work, notably in 2013, but travel times in 2016 are, if anything, more reliable and slightly lower than in previous years as there have been relatively few disruptions. The Hydro work in late 2013 took place east of this section, but the traffic backed up west of University and delayed eastbound service from about 4:00 pm onward.

For periods when the King car diverted via Queen for track construction, these times are measured on Queen Street.

The effect of construction on the Gardiner Expressway in 2014 is evident from April onward.

Removal of the stops bothways at Simcoe at the end of June 2016 appears to have slightly reduced travel times eastbound, but had little effect westbound.

• 504_University to John WB Summary
• 504_University to John WB By Year
• 504_John to University EB Summary
• 504_John to University EB By Year

John to Spadina (Updated)

This is the heart of the Entertainment District, and it can become quite congested, although not at the “standard” hours. The effect also varies depending on the weather, the day of the week and whatever special events might be taking place. Travel times westbound can be higher late in the evening than earlier, particularly when the combined effect of theatres letting out and club traffic hit King.

In 2013, this part of 504 King was hard hit by the King/Spadina reconstruction and associated diversion via Queen, and by utility construction that began just after TIFF and again later in the fall. There was some spillover effect from Gardiner Expressway construction in 2014, but the effect was greater westbound than eastbound. By 2015, travel times settled down again.

• 504_John to Spadina WB Summary
• 504_John to Spadina WB By Year
• 504_Spadina to John EB Summary
• 504_Spadina to John EB By Year

Spadina to Bathurst (Updated)

Travel times westbound to Bathurst were affected in 2014 by the Gardiner Expressway work, but mainly during the PM peak and early evening when outbound traffic is at its peak. There is little effect eastbound.

Most striking is the rise in travel times during 2016 concurrent with the 501 Queen diversion via Shaw, King and Spadina. This produced longer travel times during much of the day eastbound to Spadina where there is no transit priority signalling to move 501 cars quickly around the east-to-north turn. There is a smaller effect westbound. Some of this will be due to traffic diverted off of Queen where the westbound curb lane has been occupied for water main reconstruction.

• 504_Spadina to Bathurst WB Summary
• 504_Spadina to Bathurst WB By Year
• 504_Bathurst to Spadina EB Summary
• 504_Bathurst to Spadina EB By Year

Bathurst to Strachan (Updated)

Congestion at King and Bathurst eastbound affects travel times during the latter parts of both peak periods, and there has been a small rise in values over this segment since the Queen diversion began in 2016. Westbound times were not as strongly affected.

• 504_Bathurst to Strachan WB Summary
• 504_Bathurst to Strachan WB By Year
• 504_Strachan to Bathurst EB Summary
• 504_Strachan to Bathurst EB By Year

Strachan to Dufferin (Updated)

From Dufferin to Strachan eastbound, travel times are slightly higher in the peaks probably due to longer stop service times, not traffic congestion.

There has been a small rise in average times since the Queen diversion began in 2016 although the effect is more pronounced westbound than eastbound.

• 504_Strachan to Dufferin WB Summary
• 504_Strachan to Dufferin WB By Year
• 504_Dufferin to Strachan EB Summary
• 504_Dufferin to Strachan EB By Year

Dufferin to Jameson (Updated)

The segment westbound to Jameson is affected by the status of Gardiner Expressway construction, which accesses to it are open, and traffic signal timings. A particularly strong example of this is the early evening congestion producing much longer travel times after 6:00 pm than during the “peak” itself. This was due to traffic regulations (parking/stopping) and signal timings at Jameson where the road’s capacity was reduced before the demand actually returned to “evening” levels. There were also changes due to periods when the Jameson or Dufferin accesses to the Gardiner were closed and traffic volumes shifted between them.

The greatest spikes in travel time coincide with the CNE (late August), and the effect is becoming more pronounced year over year. The problem grows through the afternoon and evening, and is worse eastbound than westbound.

• 504_Dufferin to Jameson WB Summary
• 504_Dufferin to Jameson WB By Year
• 504_Jameson to Dufferin EB Summary
• 504_Jameson to Dufferin EB By Year

Jameson to The Queensway (Updated)

Eastbound travel times on this segment are affected by queues at Jameson for access to the Gardiner Expressway during the AM peak, and there are quite visible drops in times when that access is closed.

Westbound times are affected primarily by queues to the Queensway/Roncesvalles intersection where peak period traffic can back up several blocks to the east. The effect varies depending on conditions on the Gardiner Expressway, and it has not been as bad in recent months than in past years.

Although it is not visible in these data, day-by-day reviews show that there can be late evening congestion here associated with the outbound flow of traffic from major sporting events. Because they are one-day effects, they are submerged in average data for an entire week.

• 504_Jameson to Queensway WB Summary
• 504_Jameson to Queensway WB By Year
• 504_Queensway to Jameson EB Summary
• 504_Queensway to Jameson EB By Year