Gondolas for mass transit?

This post by Nicolas Reid was originally published on Linked in. It is republished here with permission.

Gondolas are often in the news, with manufacturers of ropeway systems proposing them as a modern option for mass transit systems in New Zealand. However, like every next big thing in transport, it’s hard to separate the marketing hype from the reality.  I recently took a trip on the Maokong Gondola in Taipei, which helped me test some of these assumptions and put some perspective on the usefulness of gondolas for public transport.

Originally used to climb mountains and span waterways, gondolas are increasingly being put forward as alternatives to typical urban transit solutions like buses and trains, even where there aren’t steep hills or valleys to traverse. This is because ropeway technologies (variously called gondolas, cable cars and aerial tramways, among other things) offer some tantalising potential benefits for public transport.

• The most visible potential comes from the fact that gondolas are suspended in the air, which suggests they can run direct from point to point, sailing over ground level traffic and congestion, and avoiding the need to repurpose street space or acquire land to build guideways.
• Another benefit is their near-infinite service frequency, with a continuous string of cabins running a headway measured in seconds rather than minutes. Catching a Gondola is more like using an elevator than waiting for a bus.
• Likewise, the constant cable loop means that a gondola trip should be fast and direct without slowing for intermediate stops, which should result in high passenger capacity. Proposals for eight- or ten-seater cabins departing every twenty or thirty seconds suggest a line capacity of several thousand people per hour.
• Finally, gondolas offer the potential for highly automated operations.  The cabins are driverless, being ‘driven’ by the cable mechanism, which suggests low staffing costs and the potential to run very high frequency departures right across the day and evening.

Back to Taipei, the Maokong Gondola is a good example of a real-world application of a ropeway system in a mass transit configuration, and on paper it offers a lot of benefits.  While it does span mountains, this Taiwanese Gondola line is more than a conventional mountain cable car. It runs from a station of the Taipei metro system to the touristy hill town of Maokong, via two intermediate stations serving a dense residential neighbourhood and a temple complex. At 4.03km length it is very long, and it features four stations in total, so it is more of a transit line than a point-to-point shuttle.

My first impression was that, while the aerial views were spectacular, the elevated route was certainly less than ideal for a transit corridor. In Taiwan, like most places, land ownership is complex and you can’t simply string a gondola line over private properties without buying the land or air rights. In practice most run over public waterways, parks or reserves, which means they’re often located on indirect alignments and less useful for urban transit than it might appear. The Maokong Gondola isn’t a straight line, rather it follows a zig zag path which seems to be based on avoiding private land and locating stations and pylons in places where it was feasible to build them. This is obvious with the main city side station, where the gondola terminus is actually 400m walk away from the connecting metro station that shares its name.

There’s no escaping the need to buy land and build multi-level station buildings and towers, which is clear in the Taipei example. This line has four stations, two further angle change stations where the cabins change cables, and 47 support pylons.  That’s a lot of infrastructure.

So what about the staffing levels? Well. despite the “automated” operation, many staff were active operating the crowded line with queues of passengers stretching out of the station and up the street. The line was very crowded for my trip, which was at peak time during a very busy holiday weekend. So, it was a good test of real-world performance under high demand like you might expect at peak times on a mass transit line.

At the end station there were two staff selling tickets, three more managing the fare gates and queues, and another two staff allocating passengers to cabins on the platform. There was a further staff member managing the lift of the multi-level station, keeping it free for wheelchair and pram users and sending most people up the stairs. Presumably there were several more staff at the other stations, and at least one in a control room somewhere. Altogether, there must have been at least a dozen people running the gondola line. So much for automation!

In terms of speed, the gondola took 22 minutes to go from end to end for my ride, which equates to 12 km/h for the 4.03km line. This is not very fast, equal to a local bus crawling in heavy traffic. In the mountain context it is undoubtedly quicker than the winding route a bus would need to follow to cross the hills and valleys, but it’s a fraction of what a surface rapid transit line usually achieves with dedicated lanes in a city environment.

I timed the headway as one cabin departing every 45 seconds, despite the long queues showing demand for more frequency. This is indeed very frequent for a transit line, but only about half the throughput of the supposed “every 20 seconds” that I see in Gondola marketing. I can only assume that crowding or other operational factors were limiting the real world frequency to 80 cabins per hour each way.

Capacity wise, each cabin has eight seats but in practice the utilisation was low. Despite the long queues, the staff sorting groups into cabins weren’t forcing passengers into every seat in every cabin. I guess there was concern about people being trapped with strangers in a small space for over twenty minutes. Groups of four or five people travelling together got their own cabin, as did anyone with a wheelchair or pram, while they did pair up groups of two or three to share, but not more. In effect the gondolas were running as private cabins, with very little public sharing of the vehicles going on. All up, I estimated the average occupancy was about four and a half people per cabin even under peak conditions, about half the theoretical capacity.

So what does this all mean for the capacity of the line? Well, during that busy peak demand period, there were running 80 gondola cabins per hour each way, and loading an average of 4.5 people on each, which means the functional maximum capacity of this Gondola line was only 360 people per hour. That’s about the same as a normal bus route running double decker bus every fifteen minutes, and a fraction of the capacity of a busway or light rail line.

It’s also far far lower than the 3,600 people per hour claimed in some proposals for urban transit gondolas. Working back on the maths, it seems these capacity claims are based on ten people in every cabin, and a cabin coming every ten seconds. Maybe this is achievable somewhere, but it certainly wasn’t being done in Taipei.

There is one thing with gondolas I could never get my head around: if they have such high frequency and run so regularly, why do I always end up waiting long queue to board one?

I think the answer to that lies in the concept of latency. While gondolas have practically no waiting time with very frequent headways, and very short dwell times at the platform, gondolas have a latency in access. They have a lag in boarding that comes from using evenly spaced vehicles of low capacity and loading them one at a time.

As an example of that, if you have 100 people at a busy bus stop, a double decker can pull up, load everyone on board and be underway again in less than 60 seconds. There’s very little lag in that equation. But if you’ve got 100 people who want to ride the on the Maokong gondola, it will take seventeen minutes to load them, four or five people at a time, into gondola cabins that come once every 45 seconds. At the connecting metro station, it’s perfectly conceivable that a hundred passengers could arrive on a single metro train, and take more than a quarter hour to transfer to the stream of gondola cars. Because they’re evenly distributed along the cable there’s not really any way to change that. Meanwhile, with a bus line it’s a trivial thing to schedule the bus to meet the metro train and transfer everyone across within a minute or two.

I’ve got two takeaways from this experience. Firstly, manufacturers’ claims of maximum capacity and speed for any flavour of transit system need to be taken with a grain of salt. If someone says X can do Y, the reply should be “show me where it actually does this”.

But overall, it’s clear that cableway transit is something that works well where it works well (like climbing mountains), but it’s not a replacement for most regular bus or rail transit routes. It’s a niche option, and that niche seems even smaller than you might think.

This post, like all our work, is brought to you by the Greater Auckland crew and made possible by generous donations from our readers and fans. If you’d like to support our work, you can join our circle of supporters here, or support us on Substack!

Share this