Tuesday, November 13, 2007

I question the effectiveness of busway

Having had the chance to observe Jakarta's famous busway, as well as discussed it with some friends, I doubt that busway is, and will be, effective.

There are several premises and assumptions in this thinking (or perhaps others have done similar approach as well). One is that the goal of having a busway system is to reduce traffic congestion on the streets where the system is implemented.

Here traffic congestion is defined as the number of cars per unit length of street with busway. If this number gets smaller with busway implemented, then it's effective. And vice versa.

There's no hidden agenda, politically or economically, in the opinion below, which concludes that installing busway lanes, indeed, increases traffic congestion.

The framework of determining the effectiveness is very much simplified, that is by looking at whether replacing a vehicle lane with a busway lane will actually reduce the overall traffic (on the other vehicle lanes).

Without considering the cost, I'm looking at the balance of demand (number of people to ride busway) and supply capacity (number of people that the busway system can transport) - on a unit of length.

Let's say a vehicle, running at a decent speed on a traffic lane, will consume an average space of 12.5 meters (4.5 meters length + 8 meters distance between front-end and back-end of two cars). That makes 80 vehicles per lane per kilometer.

If we assume 1.25 passengers per vehicle (that is 5 people for every 4 cars), that'll make 100 people transported per lane per kilometer.

The question is whether taking off one lane for busway will positively affect (reduce congestion on) other lanes. In other words, whether at least these 100 people in 80 cars will use busway.

Another assumption here is that there's no costs of switching (whether financially or based on comfort). These passengers will switch to busway (if available) once they feel there are more than 80 cars per lane per kilometer, and so on to keep the system in balance.

To complete the demand side of this equation, it is also assumed that in each bus available, 25% of the passengers are actually coming from those who do not drive. These are the people who has the buying power to switch from other means of mass transportation to busway. These people are assumed to be more experienced in using mass transportation, thus will always get in to the busway ahead of the people who are used to drive or use vehicles. In other words, these people are in higher priority.

Let's also say the average speed of a busway is 30 km/h, of which each bus will take 2 minutes to travel for 1 km. If we assume the time between two buses is 1 minute, there will be two buses at any given time in a kilometer of busway lane.

Another important parameter is maximum capacity of a bus. A friend who's a regular busway rider informed me that a normal capacity is 40 passengers, and 60 passengers in an extremely full situation ("kalau dipaksain"). Let's assume an average of 50 passengers per bus.

This will give us a capacity to transport 100 passengers (50 passengers times 2 buses) every one kilometer per one busway lane.

Since 25% of these 100 passengers, or 25 people, come from other mass transportation, only 75 people truly move from driving a car to riding a busway. These 75 people equal to 60 cars (with 1.25 people per car).

A little summary from the above rough estimate:
- number of cars "displaced" when busway lane is installed: 80 cars/km/lane
- number of "cars" that will switch to riding a bus: 60 cars/km/lane

This means, once a busway lane is installed, the other lanes will be more congested by 20 cars for every kilometer. The more lanes available, the less added congestion will occur. But it will always be more congested. E.g. if there is only one other lane, it will be more congested by 20 cars for every kilometer. If there are 20 other lanes, each lane will be more congested by 1 car for every kilometer.

Therefore, busway is not effective in reducing traffic congestion.

Imagine on Jalan Sudirman, which is about 4km long, the two express lanes (jalur cepat) will have an additional of at least 160 cars - displaced from the new installed busway lane.

With this same logic (and only with this logic), several ways to decrease traffic congestion with busway are:
- Increase the capacity of busway - either by using bigger buses (more than 50 passengers per bus) or by having more frequently buses (less than one minute between buses).
- Prioritize passengers who "give up" driving. (How?)
- Add another vehicle lane for every busway lane installed. (Then we may not need busway.)
- Create busway lanes (or other mass rapid transportations) that do not utilize regular vehicle lanes. Subway? LRT? Hmm...

I'm sure there are lots of loopholes in this exercise, mostly through the simplified assumptions (I may need to do some sensitivity analysis?). But the conclusion probably won't be too far...