I was having a trawl through recent changes to the London Underground network and noticed one change that had somewhat slipped past my attention – the adjustment of the Circle Line to something that’s now more of a “spiral route”, with the addition of a section from Hammersmith to Paddington. One of the reasons for making such a change could obviously have been the desire to get more services to Paddington, but through reading up on the change it also seems that there was a desire to shift away from the “loop” route patterns of the Circle Line.

The diagram below shows what the “Circle” Line looks like today: The difficulties in operating an orbital/loop route are noted in the Circle Line’s Wikipedia page as a reason for the change:

Orbital routes have an intrinsic problem of timetabling robustness. The trains are constantly in service and so there is little scope for “recovery time” if they are delayed. A single delay can have long-lasting knock-on effects and be much more disruptive than on a non-orbital railway. Recovery time can be created by timetabling longer stops at some stations, but this increases journey times. The current spiral route supposedly removed this problem because of the recovery time at both ends of the route.

That got me thinking about loop routes in general, and particularly Auckland’s best known example of a “loop”, the Link Bus service. I don’t catch Link buses particularly often, but when I do it is fairly plainly obvious that the buses suffer from what’s outlined above, probably to a greater degree than happened on the Circle Line because buses, when operating in mixed traffic, have many more possible causes of delay.

Link buses have an unfortunate tendency to “bunch”, because slow boarding times and unreliable traffic conditions mean that the first bus gets slowed down, while the bus behind it slowly catches up because it doesn’t have to pick up passengers and can therefore travel quite quickly. The longer the route gets, the more bunching you end up with – the end result quite often being three Link buses travelling one after the other, and then a huge gap to the next one.

Now this kind of thing can happen on any bus route, but the difference with non “loop” routes is that you generally have the opportunity to schedule a bit of “buffer” at the start and end of each run, meaning that if a bus is late (or early) there can be a re-timing of the service through the buffer, to give the best chance of starting its next run on time and therefore minimising the potential for bunching and unreliability. With the Link route (or any loop route for that matter), you are likely to have passengers travelling along each part of the Loop, (after all that’s one of their great strengths) which means that every time you have a “delay point” you annoy the heck out of passengers who have to sit at a bus stop for a few minutes twiddling their thumbs for seemingly no reason. I’m sure many readers would have experienced the “Victoria Park wait” on the Link bus over the years – to their great frustration (particularly as the bus drivers never seem to tell you how long it’s going to be).

When you’re trying to make public transport a more attractive option compared to driving, by improving its speed, reliability, convenience and so forth, having bunching buses or hugely long waits for seemingly no reason at Victoria Park don’t really help. I know of a lot of people who’ll avoid the Link bus for these reasons.

Yet I have to balance this against the advantages of “loop” routes. They do offer a connection between such a great number of places – particularly when it comes to the Link bus. Ponsonby to Newmarket: sorted; Parnell to downtown: sorted; Grafton to Ponsonby: sorted. If there are ways to minimise the unreliability, bunching and annoying waits associated with loop routes, they certainly have their many advantages.

So what might be a way forward on the issue of loops? I’m certainly not advocating for getting rid of the Link bus anytime soon – as it has been a significant success, despite its flaws. The high frequency and good connectivity of the Link has clearly struck a chord with the Auckland public right from when it was first introduced in the late 1990s. But how might we reduce the problems associated with the Link? Can we put in better bus priority measures at key points along the Link’s route where it experiences delays? Will faster boarding associated with integrated ticketing help reduce the ‘bunching’ caused by slow boarding? Can bus drivers be made aware of how far behind them, or in front of them, the other Link buses are – and adjust their driving accordingly? Is there a better place for the Link bus to pause than Victoria Park – which seems a very bus part of the route’s operation? It would be interesting to know the answers to these questions.

Furthermore, I also think that we should be wary of introducing further ‘loop’ bus routes – particularly if those routes are going to be longer than the current Link. The longer the route, the more opportunities for delays in mixed traffic, the more opportunity for bunching due to slow boarding and the more difficult it becomes to keep the route ‘on-time’ without having to introduce extraordinarily long waits at various points along the route – guaranteed to annoy the hell out of bus catchers.

So I suppose in summary I think loop routes are best if they’re kept short and if we can get as good bus priority along the route as possible. This minimises the opportunity for unreliable services and the opportunity for bunching. Furthermore, in the case of the Link perhaps we should look at having a number of “waiting points” around the route, in places where patronage is at its lowest, to allow the buses to keep on time without making passengers wait forever – as currently happens on occasion around Victoria Park.

I am curious what others think though – are Loop routes great, or are they a loopy idea (please excuse the terrible pun)?

Share this

6 comments

  1. Certainly the Circle line had a tendency to bunch trains after a problem and the bunching would last for some time.

    The linear lines tended to deal with bunching by terminating some trains early, de-training passengers, and then sending the trains in the other direction. There were certain stations where this was possible. It’d be easier to do this with the new spiral than the old loop. I think it is a good change. The only disadvantage is for people who wanted to continue around the loop past the new end of line. But most of those will find there is an easier route across the city anyway.

  2. The only reason Victoria Park is chosen as the timing stop is because it is close to the depot as far as I can make out. The Link actually comprises a number of trips, typically Britomart to Victoria is the staring sector, then Victoria to Victoria, then a closing trip Victoria to Britomart so it really is a scheduled service as much as any other.

    At one stage the separation was monitored from the depot but this is no longer done. Failing that I think there should be some sort of protocol so that if a driver catches up to the bus in front, the pax transfer and the bus that caught up waits at the stop.

    Link also needs to run clockwise Three Lamps, left at Beaumont, Fanshawe (to pick up Viaduct pax), Britomart, then on to Parnell / Newmarket. Avoid Victoria / Queen St altogether.

    1. So agree that the link should dump the Queen St dogleg. K’rd- Grafton Bridge-Hosp-Newmarket. So much better. Fit young students can walk down Symonds St and would get there quicker than a through route. Transit trying to be a car again. Stick to what do best and well.

  3. I think (two way) loops can be very effective in the right circumstances, particularly if the have the right design to allow them to be radial, crosstown and circulator routes all at the same time.

    The Sydney monorail has a system where the dwell time at each stop is managed. If your cab is catching the one in front you might be subject to a series of longer stops, say 45sec instead of 30. Such an approach still costs time and would be mote difficult to operate I bus and regular rail systems.

    Having a ‘terminus’ on the loop could be all it needs, a place where a timekeeping stop can be sheduled in. The most logical place for this is actually the busiest destination station, as this would feature the greatest proportion of people boarding and alighting and the least proportion travelling through. However this would have to be weighed up against operational considerations and systemic effects. For example Melbourne has train lines over two hours in length due to the fact they loop through the city in the middle, so a dwell stop of up to fifteen minutes is worked in at the main city station. It works for timekeeping I regards to the main peak commuter flow, but it is highly frustrating for people travelling across town or making transfers.
    For the Link, the logical thing would be to have a timing stop at Newmarket, not Victoria Park. From my experience the links basically offload and reload completely going through Newmarket, unlike Vic park which is a destination for almost nobody.

  4. One other big benefit of loop routes is that they are easy to understand or know and simple.

    Buses bunching is a proble. More buses might help as even though you may get bigger bunches hopefully more will be outside of the bunch.

    Anyway the best way to avoid bunching is better bus piority measures. Making the entire loop buslanes for the link would be great. Also I agree with Patrick Rs sugestion regarding the route. It would make the Newmarket-Ponsonby route much quicker, and if the Aotea Square area needs inclusion the route could use Wellesley St West instead of Victoria St, then go right round Mayoral Drive bfore using Wellesley St East to get to Queen St to head up to Britomart

  5. Rumour has it that the Link route is about to be changed, in conjunction with making the City Circuit a free shuttle up and down Queen St.

    My (somewhat hopeful) prediction is that they remove the uptown dogleg from the Link, and just it straight across K Rd as suggested, plus the will also skip the Victoria St detour and run the route along Fanshawe St to service Tank Farm, the viaduct and straight through Britomart.

    If there is a free and frequent bus up and down Queen St then people can easily transfer to get to the core CBD, so it makes sense to cut out those detours and leave it to be a proper circulator.

    The only real losers would be persons going to the University of Auckland campus: while there are a zillion buses on the Central Connector to connect to at either end of town its still a bit of a hassle without proper integrated ticketing.

Leave a Reply

Your email address will not be published. Required fields are marked *