Friday, 6 June 2014

Level of Public Service: D*

* Deliriously incompetent
The EIS published tables purportedly showing the projected performance of intersections in the immediate vicinity of the tram tracks when the light rail was in operation.
 We will deal with the intersections at Pitt Street and Eddy Avenue and at Eddy Avenue and Chalmers Street/ Elizabeth Street, the two most important intersections in Sydney City. The methodology used for the modelling that produced the tables is not revealed. I will quote from the EIS
"As part of the project design process localised intersection models have been developed in areas requiring analysis of traffic performance. These models were developed prior to future year traffic forecasts being available for the light rail project. As such the models use 2011 traffic volumes which enabled relative performance of design options to be assessed, but does not provide accurate future intersection statistics such as level of service to be produced.
"The modelling platform proposed includes the development of an area wide mesoscopic model with a large part modelled using a hybrid model concept."

Mesoscopic is a made-up word apparently derived from Particle Physics - TfNSW has developed its own language. It is of course gibberish.

The Level of Service calculations were in turn used to produce diagrams of degradation of traffic flow at intersections adjacent to the tram tracks when the trams are in operation to what it would be if they had not been built:

The deterioration of traffic flow of the intersections at either end of Rawson Place in the mornings are off the scale but incredibly there is no change in the performance of the Elizabeth/Chalmers Street and Foveaux Street intersection. We don't know what model of the intersection TfNSW was using - the Parsons Brinckerhoff model with one north-bound lane in Elizabeth Street or the Booz & Company model. There is no information on how the 2021 traffic flow projections were produced or how they were used to produce the LoS intersection delay calculations. There are vague references to traffic flow data from each direction but the Preferred Infrastructure Report inadvertently gave more information:

Every two or three years up until 2005 published traffic flow data on main roads. I assumed that the RTA had stopped collecting this data and was using the more sophisticated data collected for the SCATS system. In 2013 AECOM revived the practice and the traffic flow figures for one hour were published in a section of the Preferred Infrastructure Report dealing with noise and vibration impacts. The traffic flow in Chalmers Street in 2005 published in file k5sy4 was 1473 vph on the Monday morning peak and slightly less on other days and had surprisingly fallen to 1153 vph when measured by AECOM in 2013. The projected peak hourly flows in 2021 presumably produced by the "hybrid model concept" had increased to 2069 vph. Incredibly, 608 of these vehicles vanished without trace when they were transferred to Elizabeth Street northbound at Randle Street. The turn into Elizabeth Street southbound from Randle Street only takes you to Devonshire Street which we are assured can be turned into from Chalmers Street. The data that TfNSW is using the calculate road traffic noise impacts from the pedestrianisation of Chalmers Street assumes that it is not pedestrianised! TfNSW are assuming that the Assessors are imbeciles.

Infantile Paralysis

The LoS calculations were clearly produced by generating future vehicle per hour (vph) figures for the roads meeting at an intersection and then using a program to manipulate traffic light periods at the intersection. This is infantile modelling. The data has no information on what traffic is turning or travelling straight on.
Currently Chalmers Street allows four lanes of traffic to cross the Foveax Street intersection at once - one bus priority lane, two general traffic lanes and a lane for traffic to access the right turns to Cooper Street and into Albion Street. The pedestrianisation of Chalmers Street reduces the lanes crossing the intersection to two. Buses are trapped in the bus bay by traffic queued at the intersection. When kerb blisters were built in Elizabeth Street at Martin Place bus drivers refused to be trapped and routinely stopped in the next lane out reducing the general traffic to one lane. The same will happen here.

It gets worse. Randle Lane is literally a rat run. I use the lane to reach the swimming pool in Prince Albert Park and have encountered rats flattened by cars at night. Private vehicles will use the lane to avoid the congestion where the bus priority lane in Randle Street is forced to merge with a general traffic lane. The lane feeds traffic, but not buses, into the bus stop lane in Elizabeth Street.

There is a pedestrian crossing across Randle Street onto an island isolated by free-flowing traffic into Elizabeth Street southbound, but there is virtually no footpath on the western side of Elizabeth Street.

Currently traffic, including buses, crosses from Foveaux Street into Eddy Avenue two lanes at a time and there are lanes for right turns into Elizabeth Street and left turns across the pedestrian crossing. TfNSW will reduce the lanes to one - this is insane. Traffic management is not rocket science, if the number of lanes through an intersection are halved the number of vehicles crossing during a traffic light phase is halved and the length of the traffic queues is doubled.

TfNSW claims it can improve the "level of service" of the intersection by reducing the number of phases of the traffic lights from five to four. The phase eliminated is evidently the phase where pedestrian movements in parallel with traffic movements from Foveaux Street to Eddy Avenue and Elizabeth Street northbound are stopped to allow left turns from Foveaux to Elizabeth Street. Pedestrians not only face being knocked into the path of a tram by being broadsided by a lycra-shirt from either direction they will have to dodge frustrated drivers making a left turn. This is one of the most heavily used pedestrian crossings in the State.

The four phases of Eddy Avenue and Elizabeth Street intersection

Buses go to the head of the queue
Currently buses turning into Elizabeth Street can go to the head of the queue giving quick access to the heavily-used stop just south of Foveaux Street. George Street buses forcibly terminated at Coogee beach will have to join lengthened queues. The four phases of the intersection are:
Phase 1 removes left-turning vehicles from the priority bus lane allowing buses to travel untrammelled from the stop at the Devonshire Street Tunnel to the stop at Hay Street.

Phase 2 is the only phase when traffic travels from Chalmers Street to Elizabeth Street northbound. Fortunately four lanes cross the intersection at a time.
Phase 3: East-west traffic from Moore Park Road moves two lanes at a time into Eddy Avenue, mainly bound for Parramatta Road, and turn into Elizabeth Street. Pedestrians cross between Surry Hills and Central Station and Eddy Avenue.

Phase 4: Traffic to and from Parramatta Road and City Road are distributed to the eastern suburbs arterials via Wentworth Avenue and College Street.

Southbound traffic movements along Elizabeth Street take place uninteruptedly during three of the four phases but movements northbound from Chalmers Street to Elizabeth Street and westbound traffic movements along Foveaux Street only take place in one of the phases. Halving the number of lanes moving across the intersection in these phases was always going to have catastrophic consequences for traffic movements between the Eastern Suburbs and the rest of Sydney south of the Harbour. These consequences will apply seven days a week for the rest of the life of the City.

And for what? The daily lives and livelihoods and indeed the very lives of thousands of people who work and live in Sydney are being sacrificed for a shuttle between Central Station and Moore Park that will hardly ever operate during peak periods.

No comments:

Post a Comment