The Kingston Road streetcar services are often forgotten by the TTC. The service itself is highly unreliable, and for no apparent reason, this street has much worse service during weekdays than in the evening or on weekends. Yes, the riding is a shadow of its former self, but with the almost complete lack of service at times, it’s no wonder.
In the recent review of the Queen car, the TTC totally ignored the question of Kingston Road. How often should service run? Should the Downtowner and Kingston Road Tripper be combined into a single route? Is the line mismanaged, or worse, simply left to its own devices?
In this post, I will look at the service operated on the Kingston Road routes in December 2007, and I will follow up later this weekend with a review of January 2008.
December 2007 Service on Kingston Road Streetcar Routes
These links will take you to the charts discussed in the following text.
Notes About the Charts
These routes operate only on weekdays. The headway, link time and destination charts are formatted without the weekend pages seen on other route charts such as 501 Queen.
Trend lines in the headway and link time charts tend to fly off at the end of the sets of data. The reason for this is that with so few data points, the calculation assumes that the last few observations indicate the direction the trend is heading. In some cases, a few cars about to run in will be close together, and the trend line will veer sharply downward. In others, the last few cars are far apart, and the trend line swings up off the chart. This is a side effect of the calculation, and the lines should be used only as a general indication of the data’s behaviour.
The usual problems with vehicle monitoring data near terminals exist on both routes. Although departure times cannot be obtained reliably from Bingham or McCaul, headways can be measured elsewhere and round trip times to the terminals are also available. In this analysis, Woodbine has been used as the reference point in the east because it includes most of the route’s unique section. In the west, University Avenue is the reference point.
On December 24 and 31, special schedules were operated with an early start to the afternoon peak period.
For the record, the headway on the Downtowner route is supposed to be every 12 minutes in the AM peak, every 20 minutes midday, and every 15 minutes in the PM peak. The Kingston Road Tripper has the same headway during the peak period.
Analysis and Commentary
The first chart shows the actual headways at Woodbine Avenue westbound. Although the trend lines wander more or less through the points corresponding the scheduled headways, there is considerable fluctuation, particularly in the PM peak. Week 3 shows the impact of the mid-December snowstorm in the Beach with midday headways to Bingham badly disrupted on Monday the 17th, one day after the storm.
Even in week 4 which is quiet from a traffic standpoint, the scatter of headways is considerable.
The last page, showing data from all weekdays, has the alarmingly familiar “cloud” of points showing the huge range in headways encountered by riders. This is appallingly bad service.
The second chart shows headways eastbound at Yonge. The purple line for Friday the 7th swings to a very high value. As we will see later in the Destination Chart, much of the midday service never reached Yonge Street.
Week 3 is a real mess. All of the trend lines rise above 30 minutes at midday indicating that at least one car was short turned before it reached Yonge. (The typical route is via Church, Richmond and Victoria.) This is totally counterproductive when a vehicle misses the connection with the subway. It is a classic example of line management that does not care about the passengers, only about the schedule. We might think that this is the result of some horrendous congestion just before Christmas, but as we will see later, this was not what happened.
The cloud of headways is a disgrace with values close to one hour during the PM peak.
The destination chart gives one page per weekday. Each vertical line represents one car. The height of the line indicates how far it went, while the horizontal spacing gives an idea of headway variations.
December 3 is very well-behaved. We see every car going to McCaul except for the AM peak cars running in to Russell, and all of the service running in after the PM peak.
Things get worse as the week goes on. On Wednesday the 5th, a wide gap in service (nearly three hours) opens up to McCaul Loop. This appears to be caused by missing or garbled CIS data.
By the time we get to week 3, many cars don’t reach Bingham, and therefore there are fewer through trips on the destination chart that can reach McCaul. In effect, the 502s become Queen short turns rather than Kingston Road cars.
These charts show the headways and destinations outbound from Yonge Street. Note that any cars short-turned east of Yonge will not appear here even though they might actually operate to Bingham. The charts give us the service quality as seen by a rider waiting to travel to the Beach from downtown.
Again, week 3 has extremely poor service. Headways are much wider than normal and many cars only operate on Kingston Road for the purpose of turning around at Woodbine Loop.
If the problems of chaotic service arose from congestion, we would see badly scattered link times on the route. These charts show the times westbound from Woodbine and Kingston Road to University and Queen Street.
The values are fairly predictable and they are clustered around 30 minutes with an overall gradual rise from the early morning to the PM peak. The only days where there are wide swings in values are those following the December 16 snowstorm. The big rise in times on Monday, December 24th in the morning was caused by a service blockage just east of Church lasting from about 7:10 to 8:10 am. There is a comparable delay in the detailed charts for 501 Queen, but on that route many cars diverted around the area via King.
The eastbound link times show a different pattern with a marked rise in eastbound running times during the PM peak. When I review the underlying details, almost all of this is due to congestion in the core from University to east of Yonge.
The ragged headways are not explained, however, by this because the service is already irregular before the rise in eastbound running times.
Round trip times from University to McCaul Loop are fairly well behaved except in week 3. On Friday, December 7, there is a very long “round trip” caused by a car that sat in McCaul for almost an hour. During this period, all following cars short-turned at Church.
During week 3, midday service had serious problems on McCaul Street probably due to snow and parked cars.
These are interesting charts because they show that although the times vary, they do so in a quite predictable manner.
503 Kingston Road Tripper
Route 503 Kingston Road Tripper operates only in the peak period and quite infrequently. There are not enough data points to make a reasonable trend line for each day, and these have been omitted.
I am running out of words to describe hopelessly bad service. During the AM peak, the headways are vaguely clustered around the 12 minute mark, but the scatter is huge with values up to 30 minutes.
In the afternoon, the data points are smeared out over a range of up to an hour.
This is more of the same with the intriguing change that the very wide headways are seen here in both the AM and PM peaks.
The collective impression from this data is of a line where management is completely absent except in dire emergencies. Where cars are getting back on time, I can’t help wondering how many of the short turns were actually invoked by CIS Control or an on-street route supervisor. There is little evidence of regular headways arising from the short-turns.
Headways are chaotic, and the scheduled myth of a blended headway cannot possibly be achieved.
It is no wonder that riding on the 502/503 cars has dropped like a stone for the past decades.
In the next post, I will turn to the January 2008 data, and we will see how the line fared under somewhat better weather and traffic conditions.