I’ve been thinking about a number of things lately; one of them is transit to the airport. It’s a broadly popular concept, and the City of Calgary is studying some options. I was wondering what kind of ridership we would see in Calgary if we connected LRT to the airport, and what that would look like in comparison to other stations in our system.
Airport ridership statistics
What sort of ridership do cities like Calgary get at their airport stations? Here’s both Canadian cities with rail to the airport, as well the 9 cities in the broadly-defined western US with rail. I think it’s the most reasonable comparison set. It’s worth noting that virtually all the cities with rail to the airport are in bigger metros and/or have busier airports; on average, about triple the size of Calgary.
The first two columns identify the city and airport; the second two are the average weekday boardings from the rail station(s) at the airport, and an estimate of annual boardings. The passengers enplaned is the number of people getting on airplanes at the airport. The last two columns, the metro size and transit commute share, are to provide a sense of scale and context.
The coloured column is the important one. It divides the annual transit boardings by the number of passengers enplaned. This is the key; obviously a busier airport will, all things being equal, produce more passenger traffic. The boardings per passenger rate shows how much ridership can we expect, given how busy the airport is.
What’s remarkable to me is how generally consistent these values are. Here they are in a figure:
Vancouver is clearly a bit of an outlier. Dallas and Phoenix are outliers on the low side, but they’re also the areas with by far the worst transit ridership, so that’s not surprising.
The other cities? All in around the overall rate of 0.08. That includes commuter rail (Denver, Toronto), heavy rail (SF Bay) and light rail (the others). That includes metros with 24% and 18% transit commute mode share (Toronto / SF Bay respectively) and those closer to 4% (Denver, Minneapolis, Salt Lake).
I don’t know why Vancouver is such an outlier. My best guess is that because SkyTrain is an automated system, they tend to have higher frequencies in the offpeaks, which is useful for airport workers. It’s also a very good service; it goes directly to and through downtown, and might be the only service here that’s substantially faster than the car. In any case, it’s clear that basing a guess on ridership by looking at Vancouver is a very optimistic one — is Calgary really going to accomplish what 10 other cities don’t?
Estimating Calgary ridership
The general consistency in ridership rates is encouraging, because it suggests that you can ballpark airport rail ridership (given the city has moderate transit use) and have some degree of comfort. Using the rate of 0.08 boardings per passenger that’s the average, an airport station in Calgary would get around 1,800 riders per day. Here’s ridership estimates using this midpoint, as well as optimistic (Denver/Minneapolis rate) and pessimistic (Salt Lake City/Toronto rate):
Okay, so 1,300 to 2,500 passengers boarding at the airport station seems to be a reasonable guesstimate; certainly an order-of-magnitude value. Is that a lot? Here’s where that would rank amongst Calgary’s suburban LRT stations (ie all those outside the ones on 7th Ave downtown):
I left off the labels, because there’s so many stations and the names are so dang long. But the medium estimate airport station is less than half the use of the average station in the system. The zoom in below shows lower volume stations, those under 3,000 boardings per weekday — this is about 33% of the stations, but they provide only about 14% of the boardings:
The medium case has the airport between Banff Trail and 45 Street, as the 29th busiest station out of 37. The high case has it 27th and the low 33rd out of 37. Stations in this size range — Sunalta, 39th Avenue, Franklin, Banff Trail — are not high volume stations. They are worth building and the train stopping at on the way to the higher volume stations, but they aren’t worth an entire spur to themselves. And remember that the best case would be better than 9 of the 11 peer systems.
Methodology and Notes
To estimate annual boardings, I multiplied weekday average boardings by a factor of 350. If the factor was 365, it would mean that weekend boardings are exactly the same as weekday boardings; if it was 250, it would mean that nobody used the station on weekends. One strength of airport transit stations is high weekend and holiday use, which is why I used a number as high as I did.
Some places provide enough information to actually calculate this figure; in Vancouver, the scale factor is around 355, in the Bay Area it’s around 320 and in Portland it’s 368 — the airport station is actually busier on Sundays than weekdays. I used a single figure to try and get an apples-to-apples comparison as much as possible; since I went back to weekday boardings, this factor is arbitrary.
The values for airport transit are from the most recent ridership data available, as follows:
- Dallas – 2017 data including both DART & TRE
- Denver, Salt Lake City – 2017 from D. Zimny-Schmitt thesis
- Minneapolis – 2018 (both terminals combined)
- Oakland, San Francisco – 2018 average exits
- Phoenix – July 2017 to Mar 2018
- Portland – Fall 2018
- Seattle – LRT annual 2018
- Toronto – April 2019
- Vancouver – 2018 summaries
The Calgary Transit open house boards from their engagement this past summer showed Toronto as having 10,000 passengers, but this is the entire Union-Pearson Express Line. UPX was originally designed as a pure airport service; after ridership was remarkably low, the line dropped their prices substantially. It is now also ridden by people seeking an express service to downtown; from the Bloor UPX station or the adjacent Dundas West TTC subway station, it takes 22 minutes to Union Station by subway, and only 8 by UPX. Only 3,200 of the 10,000 weekday boardings are actually at Pearson.
The boards also showed higher values at Vancouver than I have here; the SkyTrain has about 8500 boardings at YVR Airport Station. However, the SkyTrain stops twice more before leaving the airport zone on Sea Island, with free service between those stations; it essentially operates as a people mover within the airport as well as a transit service to the airport. For instance, if you drove to the airport and park at the economy long-stay parking, you take the SkyTrain two stops to the terminal. These aren’t in any meaningful way transit trips to the airport, any more than Calgarians who park at the Park & Jet lot and take the shuttle bus to the terminal are. The 2018 YVR Annual Report reported that only 66% of SkyTrain trips were off-island; I’ve adjusted the boardings down by that factor.
Mean Median Mode Choice 