Stephen Rees's blog

The Canada Line last Sunday midday

The Vancouver Sun (and over the weekend the local press too) has a longish piece on how fast ridership has risen on the newest part of our rapid transit system. While generally a good news piece, there are several comments made by Ken Hardie – the Translink flak catcher – that deserve closer examination.

Hardie said that now, TransLink typically runs 14 of its 20 Canada Line trains, each with two cars, at 3.5-minute intervals, with another two trains added at rush hour.

By August of 2011, the transit authority plans to regularly run 16 trains, which will represent a 12-per-cent lift in service, every 3.33 minutes.

Hardie wouldn’t say how much this would cost.

But he noted that when “we run more transit, we spend more money.”

That is actually only true of the bus system. It is not true of a rapid transit system that has no drivers. Indeed, very early on in my career with BC Transit (as it then was) I was taken on a tour of the SkyTrain depot, and told that one of the great advantages of automatic, driverless systems was their responsiveness to unexpected surges in demand. SkyTrain staff were especially proud of the way they could deal with crowds leaving the stadium after a game. It was, they said, “just like turning on a tap”. They contrasted that with the difficulty and cost of getting operators in and paying them overtime if transit wanted to put on more service to meet a surge in demand from the PNE.

With driverless trains, there really is very little change in cost – some energy (but that is diminished by the ability of trains to produce power during braking) – and some small change in mileage based maintenance. But the system itself requires the same number of people if the trains are full or empty – and running 16 or 20 really costs very little different. What Hardie may be revealing is the impact of the P3 with InTransitBC: they may be able to charge Translink more if they run more trains. Indeed given the following – Translink is thinking about increasing bus operation (which do cost a lot more as 80% of their costs are labour ) – it may be that the secret deal is worse than anyone thought.

having buses scheduled for Brighouse shifted to Bridgeport, where commuters can catch a second, nearly empty train, from the airport.

Now that is expensive – since the schedule has to be revised (means it can’t be done until the next sheet change) and buses added, as lengthening the route increases the vehicle and operator requirement. Not exactly “turning on the tap” is it?

The newest SkyTrain cars between Stadium and Science World

For my years at Translink the biggest constraint on SkyTrain was that there were not enough trains, and there were no capital funds to expand the fleet while there was plenty of theoretical available capacity in the signalling/operating system. We could have run trains at tighter headways, or longer trains, we just ran out of equipment really quickly and could not sustain it all day.

It is possible to run trains more often on the Canada Line and there is spare capacity on the airport branch. But once again the peculiar funding arrangements with YVR seem to preclude a schedule change – for example 2 trains to Richmond for every 1 to the airport. Again that ought to be technically easy to do with the Alcatel system, but seems inconceivable on the Canada Line. It’s our money that is paying for this but we are not allowed to know the details of the contracts Translink has signed. This is not the case in other places where there are private sector contracts to run public services: just look at the amount of detail in the public domain on Britain’s privatised railway – or Transport for London’s battles with its underground system providers/maintainers – recently brought back in house due to the unsupportable demands of the private sector for more cash.

There is nothing in the story at all about what this means in the longer term. The Canada Line is capacity constrained by lengthy sections of single track and short stations: it is feasible to insert an extra car in each train but that is not anticipated for a long time. While most systems hit a physical maximum at a train every two minutes (imposed by the ability of people to get on and off trains at crowded platforms) the single track sections impose a much longer turn around time requirement.  These physical constraints have not yet bitten. It is the contractual constraints that seem to be biting now, but no-one is talking about that either.

Hardie noted he expects ridership to continue to grow, especially as municipalities continue to densify areas around the transit stations.

But all that leads to is the day when Translink does not have to subsidize In TransitBC for ridership under 100,000 per day – one of the few details of the contract we are allowed to know. What happens after that? What sort of system do we have where the demands of the operating entity are more important than the demands of the customers? Ridership won’t grow if the system is unresponsive to increasing demand. Pass ups and overcrowding do not make for happy transit riders, and they start looking at better ways to get around.

UPDATE The story now gets covered by Jeff Nagel of BC Local News. While he reads this blog, he has to talk to me to get a quote. Obviously his employers do not want to provide a link to a blogger. Selective quotation may be standard practice but given the amount of information above I think it is clear that I am aware that there is more than just power costs in train operation. It is also clear that I am pointing to the difference between price and cost – especially when it comes to P3s. But then you read me here and not in your local paper: thank you!