WestConnex and the sad reality of Level Of Service F

Before launching into some blogs on the WestConnex business case, I thought I’d go over some measurement standards that are important to practitioners in the transport and traffic engineering world. They’re important given the traffic projections in the EISs for the M4 Widening and M4 East sections of WestConnex. These ultimately have implications for the business case.

Interpreting these standards also provides the easiest entry point for residents and other interested people currently wading through the mountain of material contained in all these documents. So here goes . . .

If you know what to look for, you’ll find diagrams and tables that show something called Level Of Service (LOS) standards for different links and intersections in the road network. These show the spread and changes in road traffic congestion levels predicted by the traffic models for scenarios with and without the proposed motorway sections.

In the case of the M4 sections of WestConnex, they basically show that there is no or little improvement in congestion for many sections, and for any transport professional this sets off serious alarm bells. But you need to know how the standard works to appreciate why this is such a serious problem.

The Level Of Service (LOS) framework is used to measure the performance of both roads and public transport services using a scale that moves from A to B, C, D, E and F. A is at the end of the scale that is meant to be good, while F is at the end that’s bad and meant to indicate system failure.

The definition for each of the levels taken from the 2009 Austroads Manual gives a qualitative description for each level that goes like this:

Screenshot 2015-10-02 15.23.20

Most industrialised countries have similar manuals with the same LOS standards. The USA has the Highway Capacity Manual for example, so this framework is a common method of practice used by transport professionals all around the world. And we like using it because it is a quick way for us to gauge what’s going on with a network without having to pour over the numbers in enormous detail — although we like to do that too.

The table below shows the number-based definition for each LOS for motorway lanes. I’ve taken this from the 1988 Austroads Manual, (Part 2: roadway capacity) because its easier to follow than all the graphs in the 2009 upgrade, but they essentially say the same thing.

Screenshot 2015-10-02 15.05.39

The far left column of the table shows the LOS A through to F scale, where A is good and F is bad. To the right you can see three sets of different number-based indicators used to describe the LOS conditions for road types with design speeds of 110, 100 and 80km/h:

  • Speed: the average travel speed of vehicles in the traffic stream (note how it drops as the LOS drops from A down to F)
  • Volume to capacity ratio (v/c): the number of vehicles in the traffic stream as a proportion of the maximum possible number the road can carry
  • Maximum service flow (MSF): the maximum number of vehicles per hour that can pass a given point on the road at that LOS.

Significantly, at LOS E the maximum service flow (MSF) is 2,000 vehicles per hour.

This number is the maximum number of vehicles that can travel on the road before the traffic stream deteriorates and becomes chronically congested, or where ‘flow breakdown occurs, and queueing and delays result’ as outlined above in the qualitative description. When chronic congestion occurs the number of vehicles that can pass through the system actually drops. This number is based on observation, not theory. A good example of a road that has reached its maximum service flow, or ceiling capacity, is the Sydney Harbour Bridge.

Annual Average Daily Traffic (AADT) volumes on the Sydney Harbour Bridge and Sydney Harbour Tunnel (1975–2002)

Screenshot 2015-10-02 17.12.49

Source: Zeibots, M. 2006, Submission to the Joint Select Committee on the Cross City Tunnel by Michelle Zeibots, Institute for Sustainable Futures, University of Technology Sydney, 1 February.

In the diagram above you’ll see that in 1986 Annual Average Daily Traffic (AADT) volumes reached 180,000. From that year until 1991, traffic volumes bobbed up and down very slightly around that 180,000 AADT mark — volumes stopped the steady rate of growth that had taken place in the decades before. During peak periods, the data show that each lane with traffic moving in the peak direction carries around 2,000 vehicles per hour. Sometimes it might go over by a hundred, but never more than that and in the broader scheme of things the numbers are in keeping with the standard in the manuals for LOS E.

Now back to the LOS framework. The different sets of LOS definitions show how transport professionals make the jump from a quality-based to a number-based measurement — this is important because it tells us something about customer experience and perceived benefits. Jumping between qualitative and quantitative definitions involves some deep logical and philosophical dillemas that I won’t go into, but generally the framework does the job and helps us make decisions about whether a proposed change to a network is a good thing or a bad thing and so whether we should or shouldn’t do it.

When using the LOS framework, the joke around most traffic engineering offices — because they are places of extraordinary wit and hilarity — goes like this: Level Of Service A is really good and LOS C is not too bad, LOS E means the road is operating at its ceiling capacity and LOS F means things are really F–#ked.

Equipped with these ideas, we can now take a look at the traffic volumes in the EIS for the M4 widening and M4 Extension.

Helpfully, the authors of Appendix D which documents the traffic and transport analysis for the M4 Widening also provide LOS values for the roads and intersections surrounding the M4. These act as feeder and distribution roads for traffic entering and exiting the M4.

The following network diagram shows 2014 volume to capacity ratios for each of the road links in the network surrounding the M4 section due to be widened in Stage 1 of WestConnex. Links coloured red are LOS E or F while roads coloured green are LOS B or better and links in yellow or orange are LOS C or D.

LOS for M4 and surrounding distribution network during morning peak (2014)

Screenshot 2015-10-02 17.26.03

Source: WestConnex M4 Widening Environmental Impact Statement: Appendix D traffic and transport working paper. NSW Government, Sydney, p. 80.

See how this provides an easy way to gauge conditions for a vast network? Given that you also know that LOS E means the volume to capacity ratio of 1 has been reached and is highlighted in red, you can see the M4 is operating at levels close to its ceiling capacity similar to what could be seen on that earlier graph of road traffic volumes using the Sydney Harbour Bridge. Note how in this diagram most of the surrounding network is not too bad. We know this because many of the links are shown in green.

Significantly, this diagram cuts off the feeder roads to the start of the M4 at Strathfield. Anyone who has used this section of road knows congestion is extremely bad at this point. If it was shown, we would see a large number of roads in red around this area in both morning and evening peaks.

Now lets takes a look at the same diagram for the projected LOS for 2031 after the full WestConnex has been opened to traffic. You’ll see that there isn’t really much difference on the M4, but that many of the roads on the surrounding distribution network are coloured red and orange. This means the LOS for these roads is worse.

LOS for M4 and distribution network with full WestConnex scenario during morning peak (2031)

Screenshot 2015-10-02 17.49.47

Source: WestConnex M4 Widening Environmental Impact Statement: Appendix D traffic and transport working paper. NSW Government, Sydney, p. 139.

LOS can also be calculated for intersections and both EISs show tables with these values. Below is one of the tables that summarises LOS for intersections on arterial roads along the length of the M4 East. I’ve highlighted those that would operate at LOS F before and after the motorway sections are built.

As you can see, there’s not a lot of difference and its important to know that close inspection of these values for the M4 East suggests some key intersections have been left out. But as I pointed out earlier, if you know what to look for — the LOS ratings — like transport professionals, you can get a reasonable idea of what the outcome is likely to be.

LOS for arterial intersections in M4 corridor during morning peak with and without full WestConnex scenario (2031)

Screenshot 2015-11-21 15.12.02

Source: M4 East Environmental Impact Statement: Appendices A–G Volume 2A — Appendix G: traffic and transport assessment. NSW Government, Sydney, p. 10-9.

Now a series of roads and intersections all operating at LOS E and F doesn’t match the rhetoric of motorway advocates or the Minister. But perhaps they are talking about network conditions outside the peak periods — time periods when people don’t complain about traffic congestion.

What the traffic engineers are telling us very clearly in the two EISs for the M4 sections of WestConnex is that congestion during peak periods is pretty bad and after spending billions of dollars of taxpayer funds on building WestConnex, congestion levels in peak periods are going to be pretty bad.

Conditions like this are generally unable to provide compelling business cases for high cost road projects. We certainly saw this with the results for the EastWest Link in Melbourne. So armed with this information, it should now be a little easier to wade through the WestConnex business case and assess just how robust its claims are.