Rapid population growth has impacted the road network

The population of Perth has grown significantly, leading to increased road use. The Australian Bureau of Statistics (ABS) reports there were 1.97 million people in greater Perth in June 2013, up four per cent from 2012. Perth has been the second-fastest growing capital city in Australia after Brisbane, and the ABS estimates its population could reach 5.5 million by 2061.

Rapid population growth has resulted in more cars on the road. In 2014, the vehicle fleet in Western Australia reached 2.1 million vehicles, up nearly 315 000 vehicles, or 17 per cent, from 2009. Figure 1 details the percentage change in Perth’s population, bus and train patronage, and estimated vehicle kilometres travelled (VKT). Although there was a reported seven per cent drop in train use in 2013-14, Main Roads has not assessed its causes and impact on road traffic.

Traffic congestion is expected to get worse

There is broad agreement that traffic congestion in Perth has increased over the last decade. However, there is no single or simple measure of ‘congestion’. As a result, Main Roads uses many different measures and analyses to describe the situation, making clear comparisons or statements almost impossible. Different indicators show different performance, sometimes contradicting each other.

Common measures of congestion for peak time travel compared to off-peak travel include:

• the delay in the total time taken to complete a nominated trip
• the delay per hour travelled during a nominated period, both as a percentage and in minutes
• the decreased average speed over a nominated trip.

Main Roads monitors and reports to its executive on four nationally agreed indictors of congestion. These are travel speed, variation from speed limit (known as efficiency), variability from expected travel time (known as reliability) and a combined measure of speed and traffic ‘flow’.

Table 1 shows performance on the measured sections of the network dropped marginally in 2014 compared to 2013. This included a decrease in speed travelling towards the CBD but an increase in speed travelling away from the CBD. This is based on 78.5km of the Graham Farmer, Kwinana and Mitchell freeways. It does not include the rest of the Main Roads network.

Austroads^[1] has reported increased delays during peak periods between 2003-04 and 2012‑13. This measure calculates the extra time a nominal trip takes in peak times compared with non-peak travel. The Perth morning peak delay increased from 25 to 45 seconds per kilometre travelled. This means that a 20-kilometre trip in peak hour would take up to 15 minutes longer than in non-peak periods. By comparison, the afternoon delay increased less, from 22 to 30 seconds per kilometre, taking the delay up to 10 minutes. This echoes other measures of higher congestion on into-Perth journeys and reflects funding priorities given to ease outward congestion.

[1] The Association of Australasian Road Transport and Traffic Agencies

In 2014, Tom Tom International data showed that while Perth’s congestion has worsened over time, the city has generally held its place ‘mid-table’ among Australian capital cities. Table 2 shows Perth among most capital cities by ‘congestion index’ (the percentage increase in trip time taken compared to non-congested periods), the delay per hour travelled at peak times, and the change rate.

Until recently, Main Roads has used the Regional Operations Model (ROM) system to estimate future traffic congestion. This gives point-in-time modelling, taking data from other systems, including detailed national census data and customer surveys. The last main modelling took place in 2013, and predicted results for 2016 and 2021 against 2011 results. This predicted that the performance of the road network and traffic congestion on Perth metropolitan roads would get worse (Figure 2). In particular, it predicted that by 2021, a significant and increased proportion of the freeway and highway network during peak periods would be critically overloaded (shown in red).

During our audit, we asked Main Roads to run this modelling on the new version of the ROM system. This included extra information about public transport use, and included expected outcomes from current projects. The results (Figure 3) show a much smaller increase in congestion, and an almost total absence of critically overloaded roads. This new prediction suggests that the activities undertaken by Main Roads and others will have a positive impact on congestion. However, the new modelling also suggests that the 2011 picture was less severe shown by the 2013 modelling. We note that Main Roads is in the process of introducing a new business intelligence system called STREAMS, which will improve data coordination within the Agency.