Monday 19 March 2012

Light Rail - cautionary tales

I'd reckoned that if Christchurch City Council spent $400 million on a light rail line connecting campus to downtown, they'd need 10,000 people riding that line every day at $10 per person to cover just the capital costs if we use Treasury's recommended 8% discount rate on infrastructure expenditures; if Council's able to fund it somehow at 4%, you need 7,000 daily riders at $10 each to cover the infrastructure cost. The RedBus network carries 5.8 million passengers per year, or 16,000 per day. The draft plan suggested this route as the starting point for a larger rail system.

The December update to Council's planning documents says they'll be examining the feasibility of the  larger network. See page 110 here.
At a broadly estimated system construction and rolling stock purchase cost of around $1.5 to $1.8 billion at today’s prices (excluding ongoing operating and maintenance costs) for a staged, comprehensive city-wide network of five key routes linked to and through the Central City, a decision to initiate this project will be fundamental for the Council and equally importantly for Greater Christchurch.
Today's cautionary tale comes from Norfolk, Virginia.  They've a population of just under 250,000 for the city proper, but a total metro population of 1.6 million if we count the greater Hampton Roads area.

They put in a 7.4 mile starter light rail line for $320 million, or about NZ$400m - roughly comparable to what Council here wanted to spend on a slightly shorter line between downtown and the University. Norfolk's line, The Tide, has now been running for 6 months. They're getting an average of 4,642 riders on weekdays; they're forecasting it'll hit 7,200 daily riders within three years. They're claiming success against expectations of 2900 daily riders, but AntiPlanner puts paid to those claims; the initial projections said they'd get 10,400 riders on weekdays. And they're only charging $1.50 per adult trip. [HT: 36chambers]

Norfolk's Tide system connects a big medical centre complex, an art museum, the entertainment and commercial district, the courts, baseball stadium, Norfolk State University, and a couple of stations with park and ride facilities for mixed commuters in a metro region of 1.6 million people. It's a line roughly comparable to the Downtown to University line Council initially proposed as the start of a light rail network for Christchurch, but in a metro region that's more than four times bigger than Christchurch. And it's only forecast in three years' time to start getting the kinds of daily ridership numbers that might have a Christchurch line covering its capital costs (ignoring operating and maintenance costs).

Here's AntiPlanner on Norfolk:
Needless to say, the Norfolk line also suffered a huge cost overrun, costing about $320 million instead of the $198 million that had been predicted in 2003. It was supposed to open in 2008; instead, it opened in 2011. These problems cost the transit agency head, Norfolk city manager, and several other officials their jobs.
So it is no surprise that the current transit agency head wants people to think the light-rail line is a success. Thanks to credulous reporters whose idea of investigative journalism is to reprint transit agency press releases, many members of the public will soon forget about these cost overruns and ridership shortfalls. But it is interesting to compare the Norfolk numbers with Houston, which also has a 7.4-mile light-rail line. Only Houston’s carried more than 35,000 riders a weekday in 2010, or more than seven times the ridership of the Norfolk line. In its first year of operation, Houston’s carried more than 34,000 riders per weekday.
Or compare Norfolk with the Buffalo light rail, which is almost synonymous with “failure” partly because Buffalo’s total transit ridership–bus and light-rail together–declined from 36 million trips per year just before the light rail opened to 26 million trips per year by 2001, and ridership has hovered around that number ever since. Although it is only 6.2-miles long, the Buffalo line carried 21,500 riders per weekday in 2010, well over four times as many as Norfolk’s.
The real question is: if Hampton Roads Transit ever really thought that its light-rail line would only carry 2,900 riders per weekday in its opening year, why did they build it? 
And recall that where Norfolk's in a metro area of 1.6 million, Houston is 2.1 million for the city proper and a metro region of close to 6 million. Buffalo's greater metro population is also over a million.

I'd be putting money on a Christchurch light rail system being a worse debacle than the Dunedin stadium, but I hope I'm wrong.


  1. What the heck were they connecting to what? I spent several years in Tidewater VA, so have some local knowledge. Light rail in a tiny sliver of downtown Norfolk, and not even connecting to any military installations? Talk about a trophy project. No wonder it doesn't work.
    What are the relative population densities? I suspect that Ch'ch is denser, so more potential passengers per kilometre.
    Not that I'm suggesting the project is worthwhile...

    1. They're connecting more than the downtown-to-Uni line would connect in Christchurch. So it's not a crazy comparison for that short proposed starter line.

  2. I'd also suspect the Norfolk area was a lot less dense. Also, that area has some decent roads connecting it, whereas Chch has an almost complete absence of high speed arterial roads and freeways. While I'd agree on the basic point that Chch's light rail won't have a good return, Norfolk may not be the best example.

    1. Wikipedia says that Norfolk City is much more dense than Christchurch. The city proper (not the metro area) has population of 243k, density of 1684/km2. Christchurch has an urban density of 840/km2. The broader Norfolk metro area may be less dense, but that broader area isn't what's serviced by The Tide.

  3. I doubt we have the population base to justify such an expenditure. Especially now given the $$ needed on the post-quake rebuild work. Had we built light rail prior to the quake it would very much now be a rail to nowhere. I really hope council has the (un)common sense to shelve this project indefinitely. We are already served by a pretty decent bus network which is able to change routes depending on changing residential and business requirements, a situation we now face with whole suburbs being abandoned. Light rail would not be able to respond with the same flexibility.

  4. The only way I can see it working is with Grey Power cooperation:

    If a fifth of Rolleston’s 3,822 inhabitants are elderly, and those 765 people take seven round trips a day (approx. an 8-hr day with an hour for lunch) to and from ChCh with their Gold Cards, then the railway would pay for itself through central government transfers, maintaining Eric’s $10 pricing and the 8% discount rate.

    Assumptions: 765 people x 7 round trips (14 single trips) / day; one-way ticket price $10; all 765 eligible for Gold Cards; CCC reimbursed 100% by Ministry of Transport; NZ First enters Government at next election and stays in power in perpetuity.

    1. Note that my $10 pricing covers ONLY the up-front capital cost and none of the running costs or maintenance. And it's massively more than we're currently charging for buses.

  5. Light rail is, in essence, the same as an articulated trolley bus, on a dedicated bespoke right of way.

    If it is part of the road, it offers no advantages in terms of travel time. If it is separate, then it is useless for anything else ( a busway can always be a road or a bikeway if it doesn't work out).

    The disadvantages are considerable. Far more expensive vehicles, bigger risk of fare evasion, additional specialist maintenance for overhead lines and track.

    No one is ever willing to pay the surcharge of light rail over buses, which is around threefold. Ultra low emission buses on dedicated lanes with real time information systems would get 80% of the service quality at a third of the capital cost, and operating costs that can be fare recoverable.

    1. I can understand the case for dedicated busways. I cannot understand why those who want to subsidize public transit wish to do so so inefficiently.

  6. Please stand for Council Eric, try to talk sense to mayor and idiot CEO