Spreading the load

One of the most wasteful elements of building new roads and widening roads for peak traffic is that you end up building a level of infrastructure that is vastly beyond what a city needs for around 20 hours of the day – meaning that you have excess capacity for around 148 out of 168 hours a week. One of public transport’s great strengths is its ability to use space (whether it be road space through buses or rail space on trains) more efficiently and transport large numbers of people around during those peak times.

But this doesn’t meant that public transport is unaffected by the inefficiencies of ‘peak loadings’ either. Just as building roads for peak flows means that we end up with far more road space than is necessary for the vast majority of the time, providing high frequency peak services on public transport invariably means that during off-peak times we have a lot of empty buses and/or trains sitting around. The Northern Express service is a classic example of this – with buses running every 4 minutes at peak times compared to every 10 minutes off-peak (and that’s an excellent off-peak frequency by Auckland’s standards). With a half hour trip time between Albany and Britomart, at peak times at least eight buses in each direction are required to be on the road to keep up the frequencies. This compares to around three or four buses in each direction during off-peak times – meaning that around half of your expensive bus fleet sits around doing nothing for a good majority of the time. With modern buses costing nearly half a million dollars each, that’s a pretty inefficient use of resources.

The problem with the Northern Express is that its demand is so “peaked” that even four minute frequencies are proving to be insufficient. With plans afoot to expand the Albany Park and Ride, it seems very likely that peak frequencies will need to increase to an even greater extent in the near future, which basically means buying a whole pile more buses. Going to two minute peak frequencies would require at least 16 new buses to be purchased – with pretty much all those buses only being able to do two or three runs each day over the peak. The table below gives a rough indication of how the number of buses required to operate the Northern Busway changes quite dramatically depending on the time of day:

Now there’s not necessarily anything particularly wrong with this current situation – aside from the fact that the majority of the flash new Northern Express buses aren’t used most of the time. The Northern Express makes its biggest impact on congestion at peak times – with the busway carrying the equivalent of two motorway lanes and around one third of all the people crossing the harbour bridge at those times. In other words, if it wasn’t for the North Shore’s bus system we would need around seven lanes of traffic across the Harbour Bridge in the peak direction.

As I noted above, the problem comes when we think about how to improve the Northern Express service further. How to get even more people out of their cars and onto the flash new Northern Busway? Adding more park and ride spaces is going to generate more peak demand, which means buying more buses that will only operate a few hours a day. Going down from four minute frequencies to two minute frequencies is going to mean a huge increase in the number of buses required (effectively a doubling). That’s a lot of money to spend on new rolling stock.

(By the way, most other bus routes around Auckland are in the same position if not even worse. At least peak time Northern Express buses are full, many peak time Sandringham or New North Road buses end up very empty because they’re operating at such extremely high frequencies during peak times – meaning the marginal return on that emptyish peak service must be exceedingly poor).

So what’s the solution here? Well I think we need to get a bit smarter about encouraging people to travel during the ‘shoulder peak’ times, if it is at all possible for them to do so. Of course there will be many people – like myself – who need to be at work by a certain time of the morning and pretty much have to travel during the peak. But at the same time there are many others who, with a bit of encouragement, could quite easily travel off-peak instead of at peak times. University students often have their first lectures of the day at 10am, 11am or even later – and could choose to stay at the university until after 6pm or 7pm before catching their bus or train home, if they were given an incentive to do so.

The obvious first incentive is to have a price differential between travelling at peak times and off-peak times. At the moment a one stage cash bus fare is $1.80 at all times, but let’s say we created a price differential of making it $2.20 for peak travel and $1.50 for off-peak travel – then there would be a significant cash incentive for those to travel outside the peak window. With a smart-card ticketing system in place (as we will have in the near future) it would be a piece of cake to do.  For cash-strapped university students in particular, I think offering a cheaper travel option could prove very popular indeed – particularly as students may have the best potential to have a flexible schedule (they may stay until the PM peak has finished for studying, or specifically choose classes that start a bit later so they can take advantage of the cheaper travel.

The second incentive is to simply ensure there’s a good off-peak service available. A huge number of North Shore buses, for example, only travel during the peak hour and in the peak direction. This means that only every second run they actually do is income generating (as they don’t stop for passengers in the counter-peak direction) and once again there’s a large number of buses that spend most of their time sitting around doing nothing. Off-peak services are comparatively cheap to improve, because you generally have the buses,  the drivers, the bus parking area and so forth set up to handle peak frequencies so increasing off-peak adds a pretty small marginal cost (fuel and wages) compared to adding a peak service. At the moment, the excellent peak hour service but poor off-peak service may actually encourage people who could travel off-peak to instead commute during peak hours – as that may either be the only time their particular service operates or it may be the only time when the service runs quickly or comes at a high enough frequency to warrant using. In this respect, it becomes something of a self-perpetuating cycle of inefficiency as people are encouraged away from inter-peak services and onto peak services.

Getting back to our Northern Express example, it’s interesting to take a look at the bus requirements of two potential alternatives for improving service frequencies – one being the introduction of a two minute peak frequency and the other being the result of a deliberate effort to “spread the load” – both through better off-peak frequencies and also probably through a differentiated pricing scheme:

I would suggest that option two involves a far more efficient use of resources, and even though it would add enormously to the number of trips taken by Northern Express buses each day, it may actually turn out to be cheaper than option one – because you avoid the need to buy 16 more buses!

Interestingly, this analysis also starts to hint at the potential benefits of rail. Running electric trains during off-peak times with an integrated ticketing system in place requiring only the driver on board seems like it would have pretty low marginal costs in terms of the capacity that same train could handle at peak times. While the Northern Busway certainly works well for now, I wonder at what point the high operating costs (in terms of the enormous number of buses required to run it efficiently) will start becoming a bit of a false economy compared to rail.

Share this