|
| Action for Public Transport (N.S.W.) Inc. |
| P O Box K606 | |
| Haymarket NSW 1240 | |
| 28 February 2021 | |
Action for Public Transport (NSW) is a transport advocacy group, which has been active in Sydney since 1974. We promote the interests of beneficiaries of public transport; both passengers, and the wider community. We make the following submission on the draft document O9 - BRT/LRT Options assessment and CBA, released for public consultation in January 2021.
The draft Guidelines say they aim to promote "mode neutral" assessments, but that is not what their content would achieve.
As one of the references cited in this document notes, the introduction of the Bus Rapid Transit (BRT) concept has "again ignited a campaign that BRT can provide the same service as rail transit but at a much lower cost" (Vuchic, Stanger and Bruun 2012). Unwittingly, ATAP seems to have been drawn into this campaign.
The draft document places heavy reliance on material emanating from a core group of sources actively advocating for BRT and against rail proposals (including, but not only, light rail (LRT)). Some of the work relied on was undertaken for, or funded in whole or in part by, the bus industry or by Volvo. We do not argue that this work consequently has no useful insights and data to offer, but it is decidedly not "mode neutral".
We also note that Western Australia has received funding for the preparation of a business case that will look at the feasibility of trackless trams to connect Scarborough Beach Road with the Stirling City Centre and the Perth CBD. It seems to APTNSW that the results of the proposed trial of this system should be considered before these draft Guidelines progress any further. It is possible that this system will prove to have significant advantages over BRT systems.
Rather than entering into a debate that might soon be out of date, we would like to see ATAP focus on ensuring that assessments of proposed motorways seriously consider public transport alternatives. This is where the issue of "mode neutrality" is most pressing and of most consequence.
This document constitutes "other" guidance in the suite of ATAP assessment guidelines (indicated by its "O" prefix). As noted on p.3, the proposed guidelines would link to several existing ATAP guideline publications:
The draft Guidelines (at 4.1.2 p.38) cite a paper published by the Bus Industry Confederation (Neelagama 2014), which in turn cites Scherer and Dziekan (2012)1. They say:
It is recognized that hard factors such as travel time, cost, availability of public transport services, and car ownership have a major impact when people consider the choice between using an automobile or public transport. Nevertheless, there is evidence from the literature that rail-based public transport often is considered superior to bus systems, even in cases where quantitative hard factors are equal.Citing the same authors, Neelagama (2014) notes evidence that the preference for rail remains even when aspects such as reliability, fast travel times, and comfort are equalised.This attraction of passengers is known as a psychological rail factor, and it is used to express a higher attraction in terms of higher ridership of rail-based public transport in contrast to bus services (Axhausen et al. 2001; Megel 2001b; Ben-Akiva and Morikawa 2002; Vuchic 2005; Scherer 2010a). The existence of this rail factor is widely accepted among experts, but little evidence exists about the reasons for this phenomena.
The draft Guidelines refer (at p.39) to a study that confirms the "rail effect" in Australia:
A national 2015 study found that people in people in Sydney had a 2:1 preference for tram solutions over buses, against 2.5:1 in Melbourne, Adelaide, Brisbane and the national average, and 3.5:1 in Perth and Canberra (Hensher & Mulley 2015, p.13).
The US based Institute for Transportation and Development Policy has developed and published suggested standards for BRT. This work recognises the critical importance of factors such as dedicated corridors, and service frequency:
..., the BRT Standard (ITDP 2016 p.44) specifies that high-ranking services should operate until at least midnight and on both Saturday and Sunday as well as normal weekday services; in addition, all routes operating on the combined corridor should provide at least 4vph (15min headways) off-peak and at least 8 vph (7.5 min headways) in peak (p.70). These numbers assume multiple routes using the corridor for a higher total service and reduced waiting times; therefore, for a closed LRT system, the minimum service threshold per route to grant sufficiently short waiting times is probably closer to 8-12vph off peak / 16-24vph peak.
Service frequency is critically important to all forms of public transport (as are dedicated corridors), and it is encouraging to see this recognised more broadly. This recognition constitutes a serious challenge to the "demand-led" approach that has resulted in a paucity of bus services in many areas of Australia's cities. The draft Guidelines touch on this approach, and the alternative service-led approach, (at 4.6 on p.33). A downward patronage spiral can result from the "demand-led" approach, leading to reductions in services that accelerate patronage decline, and so on. This is a recipe for car dependency and its attendant ills.
The suggestion made in the ITDP work is that bus systems that do not meet the suggested standards do not fit the category BRT - they are not really BRT at all. In this sense, the suggested standards double as a definition.
The draft Guidelines say that they provide:
... an understanding of the essential elements of BRT and LRT, as well as the similarities and differences between them.
The "essential elements" of BRT outlined in the Guidelines appear to correspond with the proposed ITDP standards. The draft guidelines classify both BRT (of that kind) and LRT as medium capacity transit (MCT).
This is by no means an agreed definition though. Another source says:
The acronym BRT is increasingly being used as a generic term to describe any bus-based transit service with operational characteristics that improve on the speed of a traditional bus service2.
This much looser use of the term BRT would include features such as dedicated bus lanes, and bus priority at traffic lights. APTNSW strongly supports initiatives of this kind. We note (from Table 15 on p.29) that prioritisation of high occupancy vehicles has shown a 21 per cent decrease in bus travel time in Sydney.
Arguably it would also encompass cross-regional bus services like the Sydney Mxx series, which reduce the need to interchange. These services are under threat at the very time these draft Guidelines are being considered, which tends to support concerns about the permanence of high-quality bus services.
The Guidelines say that M1 Table 34 provides indicative default capital costs for BRT ('Dedicated bus lanes - dual track') and LRT ('Light rail - surface dual track') (p.51). It is not clear whether these would be appropriate for BRT at the ITDP standards - this might depend on the volumes to be accommodated.
The draft Guidelines rely heavily on sources that advocate for BRT, and suggest ways in which community preference could be shifted towards it. Some researchers suggest that the "rail factor" simply reflects the lack of passenger experience of bus systems that have the kind of attributes required by the ITDP standards. For example:
Hensher also notes that familiarity of a system impacts willingness to accept it; those with direct experience with a high-quality BRT system are more willing to prefer BRT (Balbontina et al. 2017)3.]
This argument is adopted in the draft Guidelines. They present as fact the assertion that:
"BRT and LRT can be designed to deliver similar capacities, comparable services, and overall comparable customer experience".
The claim may be plausible, but it should not be taken as empirically established. One key source (Neelagama 2014) is a report produced by the Bus Industry Confederation, based on a study tour of BRT projects in the United States. APTNSW suggests that the United States is a great place to look for cautionary tales, not public transport exemplars. It is not a country noted for excellence in public transport. That honour belongs to Japan, and some European countries.
Australia's transport fraternity unwisely followed the US model for too many years, inflicting great damage on Australia's cities and its environment. Thankfully, this has recently begun to turn around. Why, one wonders, would we do it again?
The customer preference for rail travel over bus travel is repeatedly characterised as "bias" throughout the draft Guidelines:
The ATAP options generation process requires mode-neutrality. Some have observed that this is not always the case in practice. For example, Hensher (2016) notes community bias against BRT and favouring LRT (termed the 'rail factor') (p.20)....
While there are some measurements such as customer satisfaction rankings there are also external influences which often lead to unconscious bias of preferring one mode over another (4.1.2, p.38.)
...
Additionally, historically rail-based services are more likely to provide a better passenger experience (through a combination of frequency, comfort, exclusive right-of-way and other factors), passengers are more likely to prefer tram technology (or at least light rail-shaped vehicles) over buses as a solution. This bias leads to an underinvestment in bus-based solutions, and so no new positive experiences can be generated (Mulley, Hensher & Rose 2014)(4.12.1 p.41)
...
... there is an issue of bias for rail and how this manifests in the community and the subsequent impact this has on how BRT is treated in the options generation and assessment process. Factors such as track, vehicle, and station design (impacting net trip comfort), as well as familiarity with BRT in the community, can influence the extent of this bias (p.58).
Practitioners are told to "ensure that biases towards a particular mode are not influencing how they perform project assessment" (p.39).
This advice seems unnecessary, given that it is the customers who are being accused of bias (or false consciousness), not the project assessors.
The draft Guidelines state (at p.50) that because they recognise BRT/LRT as a single public transport mode class (MCT):
... in general, equivalent treatment of costs, benefits and parameter values would apply across both BRT and LRT assessment.
They advise practitioners (4.1.3, p.21) to design and compare BRT and LRT options that:
It is of concern to APTNSW that one of the key limiting factors appears to be the frequency of services.
We are very pleased to see the draft Guidelines recognise the importance of service frequency (4.5 p.29):
Service frequency, also described as the headway between departures from a stop or passing any given node along the route, is a critical component of service levels including travel time and capacity. It determines the waiting time to access a BRT/LRT route in the first instance, and again in cases of transferring between routes. It is therefore a primary factor in both actual journey times and the perceived convenience of using the route or transferring between routes.We see the 'Turn Up And Go' (TUAG) or 'timetable free' approach as a significant advance in Australian transport thinking. The Guidelines say (at p.29) that:More frequent services have shorter wait and transfer times, and more convenient connections. High frequency routes with short headways also offer resilience against delays and unreliability from the user perspective, as the impact of a missed or cancelled departure is limited to waiting for the next service following soon after.
A defining characteristic of international BRT/LRT systems is that they will often target a 'Turn Up And Go' (TUAG) or 'timetable free' approach, where relatively high frequencies are provided at least during daylight hours. Under this arrangement a user does not need to consult a timetable or plan their travel around a particular departure. They simply go to the stop and board the next service with only a short wait. While not explicitly defined, TUAG is usually marketed at the threshold of six or more services per hour.
The NSW Future Transport 2056 Strategy differs somewhat from this conception of TUAG. The Greater Sydney Services and Infrastructure Plan (p.7) considers waiting times rather than the number of services per hour - after all, six services per hour would not be of much benefit if vehicles tend to "bunch". The Plan says:
'Turn-up-and-go' services are planned on both city-city and centre-centre corridors. This means that once customers reach their nearest main station or stop on a trunk corridor, they will not have to wait any longer than 5 minutes across the day and in the evenings.The ability to achieve frequencies of 10 minutes or less (high frequency) should in our view be seen as a "non-negotiable" for all urban public transport initiatives. If a proposed system relates to a trunk corridor and it is unable (in practice) to achieve Turn up and Go frequencies (waiting time no longer than 5 minutes), it may well be appropriate to discard it early in the options generation process.For people living within ~10km of our Metropolitan Centres as well as on local corridors, customers will have access to high frequency services (at least every 10 minutes) that will enable them to reach their nearest Strategic Centre within 30 minutes or to connect to a nearby trunk corridor, where they can continue their journey.
The instruction to compare BRT and LRT options that "Are assessed as generating the same level of patronage outcomes, or other strategic benefits such as land value uplift or property development" is particularly problematic. The draft Guidelines advise (at 5.1.1) that:
Estimations of propensities to take a particular mode need to be informed by the characteristics of the proposed mode. If comparator BRT and LRT projects both provide equivalent levels of connectivity and service quality, then they should be treated as having equivalent impacts on demand.
This statement appears to give effect to the instruction that assessors should discount "community bias". It assumes that connectivity and service quality are the only qualities that have an appreciable impact on patronage (demand)4.
The Guidelines recognise that fare elasticities are generally greater for bus than for rail i.e. a given rise in fares is likely to reduce bus service patronage more than rail service patronage. It is also noted that in-vehicle time elasticity for bus is typically lower than rail, i.e. people will accept longer rail journeys than bus journeys. The Guidelines say that:
Both of these findings are related to the typically shorter trip lengths for traditional bus services. Therefore, these findings are not relevant to BRT which differs substantially from a traditional bus service.
The basis on which ATAP comes to this conclusion is not clear. We think it is arguable. We are certainly not comfortable with the proposition that follows from that conclusion (p.47):
Given the similarities in service between BRT and LRT, all things equal, we recommend using equivalent fare, service level, and in-vehicle time elasticities for BRT and LRT in demand calculations.
Similarly, the draft Guidelines advise assessors to adopt an equal diversion rate "if the service provided under each mode option is broadly similar" (p.47). This seems to reflect a statement made in Neelagama (2014, p.92), that "Research indicates that both Light Rail and Bus Rapid Transit can have similar modal shift impacts". As noted elsewhere in the Guidelines, modal shift flows through to car parking requirements, and the incidence of road crashes (6.2.7, 6.2.8)(p.53).
The draft Guidelines refer to Table 3 in M1. That Table shows high diversion rates for Cambridgeshire Guided Busway (57% from car driver, 30% from car passenger) and the Auckland Northern Busway (Express Service) (72% from car driver, 24% from car passenger). However, it contains little information about modal shifts induced by light rail initiatives. No figures are given for former car drivers diverted to the Bundoora tram extension, although the 49% diverted from "car passenger" suggests many "serve passenger" journeys were diverted.
There are figures for only one Australian bus initiative, Melbourne Smart Bus, and that shows a combined diversion rate for car drivers and passengers of 31%, much less than the Perth Northern Railway (69% combined diversion rate from car passenger and driver, and 29% from "did not travel").
The complete picture is perhaps not yet known. If however the weight of empirical evidence is currently that "the rail effect" remains even when even when aspects such as reliability, fast travel times, and comfort are equalised5, ATAP's proposed advice is not "mode-neutral".
The draft Guidelines contain several claims that BRT offers passengers a "single seat ride", and fewer interchanges. For example:
... people without recent experience of bus travel were more likely to prefer LRT solutions, while those who regularly used buses preferred the advantage of the single-seat trip, which could mean either a preference for not having to stand during their trip, or a higher weighting of interchange penalties as a higher disincentive to use public transit (Hensher & Mulley 2015, p.15).]
These seem misconceived. For one thing, there is no guarantee in Australia that a passenger on any kind of urban passenger service (apart from a taxi or an "Uber") will get a seat at all. The draft Guidelines on p.6 say:
As a rule of thumb only, about half of all passengers on a BRT/LRT service can expect to have a seat for their trip, down to perhaps a quarter during peaks depending on specific vehicle layouts.
Whatever the mode, passengers usually prefer not to stand. From a passenger perspective, there is no doubt that standing on a bus is more uncomfortable than standing on a rail carriage of any kind. Without a handhold, it is hard for standing passengers to maintain their balance on a bus. The reason is alluded to in the Guidelines in table 4 p.14, which note that for LRT:
Ride quality is usually above a minimum threshold due to safety standards for track being more stringent than for roads. This may lead to more passengers being willing to stand for the length of their trip.
Conversely, the table notes that in the case of BRT:
Ride quality is impacted by onboard internal combustion engine and pavement quality.
In addition, the stated advantage of fewer interchanges assumes that the BRT system is an "open" system, with buses joining and departing along the route to continue their runs on the regular road system. The Guidelines note on p.27 that the customer experience can be unsatisfactory on an "open" system:
Through routes allow for single-seat rides but the number of services running through the core section of rapid transit becomes less legible for users at intermediate stations. Through routes can suffer from unreliability and bunching, as the routes operating on the trunk are also subject to traffic congestion and other delays on roads and streets beyond the dedicated corridor.
Consequently, it is said that (4.4.2 p.26):
... once a city's population grows above a certain threshold 'Closed' systems will be required because the sheer volume of passengers will demand customised solutions that provide higher frequency of services with reduced cross-compatibility.
The Guidelines contain a lengthy discussion of various studies of the impact of transit modes on land use value, beginning on p.42. They state:
... there is some contention around the impact of BRT on land value compared to LRTs or other rail- based modes. Lower values are sometimes attributed to BRT's perceived lack of permanence compared to rail-based infrastructure. For example, the ACT Light Rail Business Case states:
"A quantifiable increase in residential and commercial property values has been demonstrated in areas in close proximity to light rail alignments. The same increase in land value does not occur from new bus routes. This is partly because of the permanence of light rail systems versus the relative flexibility of changing bus routes..." (Capital Metro Authority 2014)
They argue that:
There is unlikely to be much disagrrement (sic) that LRT would be seen as more permanent that (sic) regular on-road street bus services in mixed traffic. On the other hand, a comparison with BRT in a dedicated right-of-way (rather than street bus services) is quite different ... In principle, a BRT can be designed with most of the characteristics of LRT. Furthermore, a perception of impermanence can be dispelled over time, with higher land value uplifts found for new BRT projects within cities with already established BRT systems (Zhang 2019).Railed modes such as LRT are generally seen as providing better land development outcomes than tyred BRT systems. Hypotheses as to why this is the case have included the perception of route permanency, aesthetic qualities and level of service provision, all of which can be applied at least somewhat to BRT systems but need to be quantified in further research (Hensher 2007). Other claimed elements might include pollution and noise, which make BRT stations less pleasant places to be at than LRT, but these may be resolved in the near future with technological advances and the application of electric buses.
In any event, the draft Guidelines note that under the existing ATAP Guidelines Part O8, land value uplifts cannot normally be treated as benefits in CBAs of transport initiatives "because to do so would double count user benefits capitalised into land values". They say that: "The discussion in this section is commentary about the relativities between BRT and LRT in relation to land use impacts, not guidance for benefit estimation". Since it seems little can be said with confidence at this point about the impact of BRT on land values, we are left wondering why ATAP feels it necessary to provide this commentary at all.
The draft Guidelines note that ATAP already provides default parameter values for in-vehicle time, travel convenience, and vehicle and station quality in the M1 Guidelines (6.2.3). Parameter values are typically split by modes of Rail, Tram, Bus, and Ferry. The draft Guidelines advise practitioners to use the 'User benefit parameter values' for Trams in the case of both BRT and LRT projects:
BRT, when delivered adequately, provides an emulation of the LRT user experience including in terms of onboard and station comfort (Neelagama 2014, p.31). Given this, it is not typically appropriate to use Bus parameter values for BRT projects for user benefit calculations, as Bus values are more reflective of street buses.If a project is branded as BRT (or LRT) but does not meet the minimum infrastructure and service standards assumed for BRT/LRT then differing parameter values for consumer surplus inputs should be used.
It would seem that if a BRT project does not meet the required standard, it is a bus project, and the parameter values for buses should logically be applied.
The Guidelines (p.54) contain a discussion of the environmental externalities associated with BRT and LRT projects. These include air pollution, noise pollution, water pollution, greenhouse gas emissions, impact on nature and landscape, and urban separation:
Compared to a BRT powered by internal combustion engine (ICE), generally, LRT produces less noise (though noise for travelling around sharp curves can be loud) and less vibration and no local air pollution (Vuchic et al. 2012). For both BRT and LRT projects engineering and design measures can be taken to substantially mitigate noise impacts (U.S. Department of Transportation Federal Transit Administration 2016). Therefore, the precise design of a BRT/LRT project will determine the extent of these externality impacts.BRT and LRT projects can differ in their greenhouse gas emissions. LRT is typically assessed as creating substantially fewer carbon emissions per passenger kilometre than a traditional bus (Puchalsky 2005). However, this performance is dependent on a range of factors including the mix of energy sources powering LRT and choice of bus engine technology.
The above discussion illustrates that externalities of BRT and LRT can differ but are highly sensitive to the context and how a project is designed and implemented. Accordingly, while default parameter values can provide indicative values for externalities, the particulars of a BRT/LRT project need to be considered in fully assessing environmental externalities.
Transport is one of Australia's top emitters of greenhouse gases. APTNSW believes that the introduction of new diesel buses into Australia's bus fleet (whether for BRT projects or otherwise) should be avoided from now on. Alternative technologies are now readily available, and time is running out.
APTNSW thanks ATAP for the opportunity to comment on this document. Unfortunately, it is hard to escape the conclusion that ATAP has somehow been drawn into a campaign to promote Bus Rapid Transit. We do not think this is consistent with ATAP's intention to consider proposals for BRT and LRT in a 'mode-neutral' manner. Moreover, we are aware that a trial of an electric vehicle described as a "Trackless Tram" will commence shortly in Perth, which may render the old debate about the relative merits of BRT and LRT obsolete.
With due respect, we suggest that the draft Guidelines should not proceed further at this time.