Recent comments I've read about service quality and rolling stock shortages due to vehicle failures on the light rail on this and other forums lead me to look at the reliability statistics again. Hitherto, without looking too closely, I've seen that light rail reliability has been, on the surface, reasonably good - better than buses and suburban trains, but not up to the level of metro and ferries.
However, a closer look reveals some surprising qualifications on this.
https://www.transport.nsw.gov.au/data-and-research/passenger-travel/sydney-light-rail-performance-reportsLike other modes except metro, which is held to an exact standard, light rail gets a statistical leg-up with a 4 minute tolerance on time and the general target is 97% of services running on time within that definition. Not much of a standard for a mode that runs in its own exclusive lanes, but there we go.
Within those parameters, we see that:
L1 consistently fails to meet its on-time running targets, though I figure that it's "saved" from contractual questioning by the 4 minute tolerance. This is a line that, not only has its own street lanes, but is a grade-separated "railway" for the majority of its length and has a generous stipulated journey time of 37 to 38 minutes, which is about 5 minutes more than the journey should need for that length and number of stops. Ye gods, what does it take to lose punctuality in that operating environment? I think it takes Transdev, which has been operating L1 in the same lackadaisical way for 25 years.
L2/L3, on the other hand, has consistent 99% punctuality, which sounds absolutely marvellous until you read the fine print that shows that it has been allocated 37-40 minutes to complete the journey, which is 10-15 minutes longer than the journey should take if it is run competently (indeed the timetable says about 32 minutes). Quite obvious shifting of the goalposts by TfNSW there, to protect the contractor, not the customers whose interests they are supposed to be acting in.
L4 is a good laugh. It started off with about 98% reliability and then in August 2025 reliability dropped off a cliff. Turn to the small print and we see shifting goalposts again. Until April 2025, it was given 35 minutes to complete the journey, then till July 32 minutes and then from July 29 minutes, a trip easy to accomplish for the length, number of stops and part of it being grade-separated "railway". This time, in fairness, TfNSW shifted the goalposts to raise the performance expectation, favouring the customer, but the operator hasn't been able to follow through.
It's impossible to get internal information about what's happening, but from the rate of vehicle downtime, an engineering assessment might be that they're finding that those fixed bogie trams are taking a costly toll on both themselves and the tracks and have to be slowed right down on curves (of which there are plenty on L4), thus extending journey times. There could be other competence issues of course, we don't know.
Generally, a potentially great system, let down by poor execution. It hasn't done L2/L3 patronage any harm though as it's booming. L1 patronage still hasn't fully recovered after covid and Parramatta is looking encouraging, but too early to reach firm conclusions, especially as some supporting projects in Parramatta, notably the metro, are yet to be completed. Above all, we should bear in mind that the decision to replace bus routes with light rail is still a good one as the trams routes carry much more patronage and have much more capacity than pretty much any bus route in Sydney.
Tony P