Myth: The purpose of tram priority is to stop trams running late

Myth: The purpose of tram priority is to stop trams running late

Myth: The purpose of tram priority is to stop trams running late
Fact: Tram priority can and should be used to increase the speed of trams, thus providing both reduced travel times and increased frequency with the same number of vehicles and drivers.

One of the more common complaints about trams is that they are slow. Trams are held up by queueing cars, turning cars and cars that unlawfully double-park or block intersections, and are disrupted by traffic lights operating on cycles that favour cars. Until recently, little attempt was made to help trams get through intersections more quickly, mainly because the road lobby opposed – and still opposes, to some extent – any measures that might speed up trams at the expense of cars.

In recent years, the traffic engineers have consented to the insertion of tram-only phases in traffic light sequences: the little white ‘T’ lights that come on for a few seconds at a time. Unfortunately, while in some places this tram-only phase has improved things, in other places it’s just made things worse – because the tram is often prevented from moving except when it gets its special signal. Thus in many locations, tram-only phases are used deliberately to allow cars in lanes adjacent to the tram to proceed before the tram – the opposite of proper tram priority. This happens, for example, at the corner of Lygon Street and Brunswick Road, where trams stuck at the red light are delayed an additional 30 seconds to allow adjacent cars to turn right.

The upshot is that even today, tram travellers in Melbourne face delays greater than those faced by peak-hour motorists, even when the time to stop for passengers isn’t counted. Research evidence indicates that trams are delayed on average by over two minutes per kilometre in central Melbourne, and by over one minute per kilometre in the suburbs, due to factors other than passenger boarding (mainly traffic lights). This compares with ‘unacceptable’ delays to motorists of 49 seconds per kilometre in peak hour reported by VicRoads.

In other cities that have trams, the trams are given some measure of priority over cars, for example by interrupting traffic light sequences to allow trams straight through intersections, and by measures that discourage long queues of traffic from forming on narrow streets that carry trams. After all, it takes only five seconds to move a tram carrying 100 people through an intersection, even if it takes a full minute and a half to get the same number of people through in cars. There has been some hope that we may at last see Melbourne begin to catch up with cities like Zurich, Toronto or San Francisco, with the establishment of the ‘Think Tram’ priority programme in 2004.

Unfortunately there are still some very wrong ideas circulating about the purpose of and supposed impediments to tram priority, which means that even the generous Think Tram funding is not being spent in a way that is helpful to Melbourne. Other pages debunk the common notions that speeding up trams requires tram stops to be moved, or even eliminated.

There is also a view within the planning bureaucracy that tram priority is simply a remedial measure, to be employed only when trams are running late and need help to catch up to the timetable. Thus, newspapers report traffic engineers as saying that when installing special priority signals for trams, special attention must be given to ensuring that they don’t trigger in the (unlikely) event that the tram is running early. It’s as though these engineers think the biggest problem with tram operations is that trams are running too fast!

The same thinking is being applied to bus priority. The 2004 upgrade to the Warrigal Road bus route included technology to insert a special phase into traffic signals to allow buses through quickly, rather like the current hit-and-miss approach for trams. However, on Warrigal Road the special phase is not automatically triggered by a transponder, as is the case elsewhere. Instead, a bus that is running late (and only when running late) sends a signal to a central Vicroads computer, and the computer triggers the signals. Similarly, bus priority measures implemented in Sydney only operate when buses are running more than two minutes late.

The sluggishness of Melbourne’s trams and buses is built into current timetables, which include plenty of slack (typically one-third of the running time) to allow for delays due to red lights and traffic queues. From time to time even the slack allowed in the timetable isn’t enough, and trams or buses run late. If priority measures are introduced only to deal with the most serious delays and maintain the current timetables more reliably, that would only improve services to the lacklustre standard they were at in the 1980s. Our tram services would still be among the slowest in the Western world; our buses would continue to be a third-rate alternative to car travel.

If tram priority can be used to allow a tram to catch up a few minutes along its route if it is running late, why not use it more effectively to make an on-time tram a few minutes faster? This is what effective tram priority means: once priority measures allow trams to travel faster, the timetables themselves can be speeded up without sacrificing reliability. Once trams are timetabled to run faster, more people would be encouraged to use them, reducing traffic that once held them up, leading to even more gains. The improved running times can also be used to improve frequency: for example, saving 10 minutes in every hour of travel time allows an increase in frequency from 12 to 10 minutes without any more rolling stock or drivers. This would improve patronage even further.

Effective tram priority is called on to do its job all the time, not just in emergencies when trams run especially late. A faster timetable means a faster and more frequent service can be provided with the same numbers of vehicles and drivers. It can even run on time, as Zurich’s system does.

Return to index

Last modified: 22 September 2009