The Office of Road Safety commissioned Professor Max Cameron to provide expert policy advice, including statistical analysis and modelling, to assist in the determination of the appropriate placement of fixed spot-speed and speed/red light camera technology in Western Australia. Professor Cameron had previously provided detailed analysis and modelling in two reports, Development of Strategies for Best Practice in Speed Enforcement in Western Australia (2006) and his further Supplementary Report (2008). This earlier analysis had not considered the specific sites to be given priority for fixed spot-speed cameras, and had not analysed the benefit-cost ratios (BCR?s) and priority locations for speed/red-light cameras. The specific tasks required were: Apply existing model to calculate BCR?s for intersection and other potential fixed camera sites to guide the effective and efficient placement of fixed spot-speed and speed/red light camera technology; Present the results of this modelling exercise, including recommendations; Provide expert advice to assist in the placement of fixed and speed/red light cameras in Western Australia; and Report in writing on the benefit/cost ratios for cameras at particular locations taking into account injury severity, crash types, traffic volumes, offence detection rates, and capital costs for development. The method closely followed the analysis in the previous reports, in particular the analysis of road links to be given priority for the installation of point-to-point average-speed camera systems in the Supplementary Report. The economic benefits from the road trauma savings were calculated using the latest estimates of the social costs for crashes at each crash severity level, published by BITRE (2009), updated from 2006 values to 2009 using the Consumer Price Index, as follows: Fatal crash $ 2,879,368 Severe injury crash resulting in hospitalisation $ 288,776 Injury crash resulting in other medical treatment $ 16,001 These unit social costs are substantially different from those used in the previous reports. Fatal crash savings are valued 25% higher than previously, but hospitalisation crashes are valued almost 50% lower. The cost of processing the detected offences and issuing each offence notice was re-examined to clarify a misinterpretation of the data requested previously. A substantially higher unit cost than that used in the previous reports was found to be more realistic. For the analysis described here, this unit cost was inflated to 2009 values using the Consumer Price Index to provide an estimated processing cost of $6.04 per offence notice issued. Fixed spot-speed cameras Based on experience with fixed spot-speed cameras in NSW and UK, it was assumed that each of the fixed camera installations would achieve 15.52% reduction in serious casualty crashes (those resulting in death or hospitalisation) and 7.76% reduction in medically-treated injury crashes. The savings in social costs from these reductions in annual crashes on each road link were calculated by applying the unit social costs of crashes to the expected savings in serious and non-serious casualty crashes, respectively. Table I summarises the economic analysis for the Perth freeways and other major road links in categories by length. The detailed economic analysis of the major road links indicated nine links with BCR greater than one. There were no major road links less the two kilometres long on which a fixed spot-speed camera would be economically justified. The longer road links on which fixed spot-speed cameras are economically justified are predominantly those 5-10 kilometres long, thus having larger potential savings in crashes and social costs to off-set the capital investment. Alternatives to operating the fixed spot-speed cameras on Perth freeways continuously were examined. Based on experience from Sweden, suggesting no less effectiveness if cameras were operated only intermittently, it was assumed that 10,000 speeding tickets per month would be issued for speeding offences detected by these cameras. The cost of speeding ticket processing would be reduced substantially and the BCR almost tripled (Table II). Another alternative to fixed spot-speed cameras on Perth freeways would be point-to-point average-speed camera systems. This scenario has been previously analysed in the Supplementary Report, however that analysis is not directly comparable with the economic analysis of fixed spot-speed cameras in the same road environment shown in Tables I and II. This is because the previous analysis was based on the earlier social costs of crashes and lower unit cost for processing each speeding ticket issued. The previous analysis was adjusted to reflect the same crash social costs and infringement processing costs as used here (Table II). The capital investment to provide the point-to-point camera systems was estimated to be $4.9 million compared with $3.1 million for fixed camera systems, however it was estimated that $10.44 million in social cost savings per annum would be achieved from point-to-point cameras compared with $4.68 million from fixed spot-speed cameras. No reliable crash-based evaluation of speed/red-light (SRL) cameras has been found. The estimated effect of these cameras was derived from the well-documented 25-30% reduction in casualty crashes achieved by red-light cameras and assuming that the crash reduction effects of fixed spot-speed cameras would apply to the remaining casualty crashes not eliminated by the red-light camera mode of operation. It was assumed that casualty crashes would be reduced by 25% due to the red-light camera component, and that of the remaining 75% of casualty crashes, the serious casualty crashes would be reduced by 15.52% and the non-serious casualty crashes by 7.76%. Sixty of the 860 signalised intersections (7%) currently have red-light camera infrastructure, through which the 30 wet-film red-light cameras have been rotated for many years. Casualty crashes at these intersections are expected to be 25-30% lower than would have been expected if no red-light camera had been installed at each of them. Their casualty crash rates were treated as being 33.3% higher to off-set a 25% reduction due to red-light camera installation. This had the effect of increasing the expected benefits due to SRL installation by 33.3%. MRWA were unable to provide data on the entering traffic volumes at all signalised intersections, but were requested to provide estimates of the entering traffic volumes at each of the top 60 signalised intersections, ranked in order of the expected savings in social costs of crashes if an SRL was installed. Four intersections that currently have red-light camera infrastructure were ranked in these top 60. The maximum traffic volume per leg was used to estimate the entering traffic volume which would be monitored by an SRL. The expected number of detected speeding and red-light offences at each specific intersection was estimated by applying the offence rates found during the trial of an SRL at Riverside Drive and Barrack Street, Perth, during 2009. Fourteen of the top ranking intersections have either had full-control right turn signals (FCRTS) recently installed or they were planned to be installed in the near future. It was noted that FCRTS could erode the expected benefits of an SRL by reducing the severe casualty crashes before SRL installation. To allow for the effect of prior installation of FCRTS at the 14 intersections, their casualty crash rates during 2005-2009 were reduced by 26.6% and their serious casualty crash rates by 27.4%, before a revised economic analysis of the benefits of installing SRLs was calculated. Table III summarises the economic analysis for the top 17 signalised intersections ranked by estimated BCR if an SRL was installed. Included are two intersections that currently have red-light camera (RLC) infrastructure. The expected higher savings in crash social costs and greater BCRs from SRL installation at these intersections are shown in brackets. Also shown in Table III is the economic analysis of the existing SRL at the intersection of Riverside Drive and Barrack Street. This analysis was possible because the numbers of speeding and red-light offences detected per day were known from the trial during 2009. The potential saving in social costs of crashes at this intersection is relatively low and hence so is the estimated BCR for SRL installation. The locations of the twelve existing red-light cameras in WA to be replaced by SRLs have been announced. An additional 18 signalised intersections need to be identified for SRL installation, one being the existing SRL at the intersection of Riverside Drive and Barrack Street. If BCR of SRL installation is to be the principal criterion for selecting the remaining intersections, then the other 17 intersections listed in Table III would be appropriate. Additional economically-warranted intersections are listed in the body of this report. (Author/publisher)
Abstract