Back in Part 2, I wrote about the proportion of service that actually arrived at the terminals in the Beach, High Park and Lake Shore.  A comment from a reader (actually in the Part 3 thread on headways) suggested that I include trend lines in the charts for this topic, and I linked a version from my reply.

Now, having caught my breath from all those weekend postings, I offer a few comments on the new charts.

Proportion of Service Reaching Terminals on 501 Queen

Page 1 shows the proportion of cars reaching Neville on weekdays from December 3 to January 31 broken down by three-hour time period.  Trend lines are included to show the overall patterns.

Generally speaking, the worst period is the afternoon peak (1500 to 1800) with midday (1200 to 1500) and early evening (1800 to 2100) roughly tied for second last place.  The best service, stated as a percentage of total trips leaving Yonge eastbound, occurs after 2100.

Seasonal fluctuations and the impact of the mid-December snowstorm are evident here.

Page 2 shows the proportion of service reaching Humber.  This chart starts off better than the Neville plots for December, but there is a clear change in late January when the proportion of service getting to Humber goes down.  This suggests a change in line management strategy.

Page 3 shows the proportion of service reaching Kipling.  I chose this point to put the best possible face on service to southern Etobicoke, and Kipling short turns are included here.  Note that the scale is different here than on previous charts because only half of the service is supposed to get past Humber under ideal conditions.  Values above 50% occur because the numbers are relative to the total cars leaving Yonge Street westbound.

Note that the pm peak is consistently at about the 40% mark.  Relative to the expected 50%, this means that about 20% of the service destined for the Lake Shore never gets there between 1500 and 1800.

Page 4 shows the situation at Long Branch, and you can see the effect of Kipling short turns.  On this chart, the pm peak trend line rides between 30% and 40% meaning that there are days when over one-third of the scheduled service never reaches Long Branch Loop.  On an 11-minute headway, this produces unacceptably irregular service.

The late January decline noted above in the Humber chart is echoed at the points further west.

Page 5 shows the ratio between the values for Humber and Neville.  If short turns are affecting both ends of the line equally, the trend line should sit at about 1.  A mid-December rise in the values from 0900 to 1800 shows that more cars were short turned in the east end.  This was probably a combination of traffic congestion eastbound to downtown (as discussed in other articles) and snow delays in the Beach.  In late January, the situation changes, and it is the west end that has more short turns in the period from 1500 to 2100.

Page 6 shows the ratios at Long Branch.  If the level of short turns were equal, proportionately, the trend lines should sit at 50%.  Although they are clustered around this value, there are clear differences by time of day.  The large rise in the midday ratio (yellow) corresponds to the aftereffects of the December storm in the Beach.

Looking at these charts, it is important to see not just the trend lines, but the considerable day-to-day fluctations in values.  The trend lines show that there are consistent patterns over the two-month period, but there are some very wide swings in some of the individual data points.  For example, on one day, only 40% of the pm peak service actually reached Long Branch.

In previous articles about the Queen car, I have mentioned the difficulty of getting reliable information about the length of layovers at the terminals.  However, there is an indirect way to do this, subject to certain caveats:  use the round trip time from somewhere near the terminal where data is “well behaved”.  As long as the travel times to and from the terminals are fairly constant, this will give us a view into how long runs actually sit at the end of the line.

The caveat is important for Neville Loop because the section from Woodbine (the next timepoint used in my analysis) and Neville can be affected by snow and congestion delays.  However, we do have fairly good data for the westbound trip (see Part 4), and can use this to temper ourview of the round trip times.

At Long Branch, congestion is not an issue, and we can use the variations in round trip times from Kipling as a surrogate for terminal layovers.

Round Trip From Woodbine to Neville — December — January

December saw a lot of congestion at various times in the Beach both due to shopping traffic and to snow on December 16.   Although the charts don’t reveal as much as I would like about layovers, one thing does stand out — the general increase in round trip times during midday and in the evening.

The midday increase, especially on Saturday, is likely shopping-related, while the evening increase is almost certainly due to longer layovers.

In January, the data values are more tightly clustered, and the midday peak has almost completely disappeared.

Round Trip From Kipling to Long Branch — December — January

In the west end, things are somewhat different, and the data values show quite a spread from the trend line.  This is almost certainly due to long layovers at Long Branch Loop.  The effect is quite strong in week 4 (after Christmas) when the line as a whole is less congested.

December Saturdays are all over the map.  On December 1, the line operated with buses until about 1800 and they tended to handle the Kipling to Long Branch trip faster than the typical streetcar times, probably because they didn’t sit long at the loop.  (Their trip times to Humber, which can be seen in Part 4, were longer, however.)  The very wide spread in times shows that there was considerable room for layovers at Long Branch on Saturdays for many runs, and under this situation the problem of on time departures from the terminal can become an issue.  I will turn to that subject in a later installment.

Sundays in December show a similar pattern to Saturdays, but with less spread.  Christmas Day is particularly notable for generous layovers.

The pattern for January is similar to December.

As an operational issue, I don’t object to giving staff a chance for a breather, particularly when the trip from Neville can take as much as two hours on a very bad day.  However, the combination of generous layover times, long routes, and route segments (e.g. Yonge to Roncesvalles) where congestion can double the running times makes is a recipe for poor service.

Long “recovery times” encourage operators to take them as layovers whether the time is available or not, and the variations in the scheduled times shows that this is more a requirement of blending service at Humber than of providing a standard length layover.

If Queen is broken into shorter routes, operators will have more frequent chances for a break because they will be at terminals more often, and they may be more willing to shave time from a terminal layover when they are slightly late.

In the next article, I will turn to the question of headway reliability during the am peak.

In previous posts in this series, we have seen where cars on 501 Queen actually go, as well as the gaps and unpredictability of service both downtown and at the outer ends of the line.  Now I will turn to the length of time cars take getting from one place on the route to another, and how this varies both over the course of the day, and from day to day through December and January 2007/08.

We hear a lot about traffic congestion and the need for better transit priority.  If such a scheme is to benefit riders as a whole, it must address the locations and times when streetcar service is slow.  Often this is not the “obvious” time or place– the peak period, downtown — and priority schemes focussed on this narrow time and location will do little to improve service.

This article contains a series of charts that are similar to the headway charts in Part 3.  Data are organized into groups by week (for weekdays), Saturdays and Sundays/Holidays.  Instead of headways, the times shown are the intervals between a car’s appearance at two locations. 

When these times are unvaried and show little scatter, then there is no congestion or variable delay due for stop service, and almost no opportunity to change running times.

When these times vary a lot, but in a predictable way (moving up and down over regular times each day), this shows regular variations in traffic levels and stop service time.  Delays caused by traffic signals can be addressed through priority schemes.  Delays caused by stop service can be address by increased use of all-door loading.  Delays caused by congestion, especially those outside the peak, can be address by traffic restrictions on parking and turns at intersections.  These will not be popular in neighbourhoods outside of the core where the main streets are important local commercial strips and the streets are the grid through which drivers access the residential side-streets.

Where these times show unpredictable spikes or move away from a regular pattern, this is the result of some event like a storm, a traffic accident (possibly on a nearby street such as the Gardiner Expressway with a spillover effect), or an unusual rise in traffic (for example around the club district or on New Year’s Eve). Continue reading →

The 501 Queen car is a mess — that’s no surprise to anyone.  In the past two posts, I have looked at where the cars go.  More, to the point, I have looked at where they don’t go thanks mainly to short- turns.

Riders are frustrated not just by the unpredictable destinations of Queen cars, but by the lack of reliable headways.  The schedule may say there’s a car every 6 minutes, but this is at best a general average.  TTC “on time” goals say that an error or plus-or-minus 3 minutes is an allowable margin, and this means that gaps of up to 9 minutes, followed by only 3 minutes , is “on time” performance. 

This ignores the fact that the gap car will carry 3 times the headway, be much more heavily loaded, and the average rider’s perception of the service will be much worse than the average implied by the schedule and the loading standards.

If the service actually stayed within 3 minutes of the advertised schedule, life would be annoying, but tolerable.  There might even be a car in sight much of the time, if not the “always” of the TTC’s slogan for good service from the 1920s.  We should be so lucky. Continue reading →

In the previous post, I presented the information about the destinations of cars on 501 Queen with details over the month.  One issue right at the end of that post was the eternal east-vs-west debate:

Does the Beach get more service than High Park?  Are more cars short-turned in the west end than the east end to keep the Beachers happy?

Although there appears to be some indication this might have developed in late January, overall the level of short-turning is quite bad at both ends of the line, usually in equal measure. Continue reading →

In anticipation of a TTC report on measures to improve service on the Queen car, there will be a series of posts here over the Victoria Day weekend.  These will review data for December 2007 and January 2008 on 501 Queen and related routes.

I am going to take a different approach in the sequence of articles from earlier rounds.  In those analyses, I began with the graphic timetables of route operations showing first Christmas Day 2006 and then moving on to other days of interest.  Later came Headway and Link Time analyses for individual days, and later for the month as a whole.  Finally came charts showing the short-turning and service reliability as seen outbound from Yonge Street.

In practice, now that this entire process is fairly well understood (at least by me), I am going to turn things around and work “backwards” from the monthly charts to specific days.  The monthly summaries reveal days and locations where interesting events or practices might be found, and of course they also show the overall pattern of transit service.

I have made some changes in the charts both to make the headings “friendlier”, and I plan to standardize on the filenames I used for the linked PDFs.  (Really!  I promise!)  Some of the changes that readers have asked for will not be found in this version such as expanding the daily graphs to have more per day (and hence fewer hours per page).  This is a tradeoff between having charts that are too busy and spreading peaks over multiple pages.  Also, I have deliberately kept the pages in letter size format so that people can print them without having legal sized paper.

For this first set of comments, here are the charts of vehicle destinations outbound from Yonge Street.

December 2007 Westbound to Humber and Long Branch

December 2007 Eastbound to Neville

January 2008 Westbound to Humber and Long Branch

January 2008 Eastbound to Neville

These take some explaining, but they provide a lot of information. Continue reading →