North West Rail Link

Single Deck and Double Deck trains compared


Sydney's Rail Future proposes a new generation of single deck trains for which significant travel time savings and line capacity benefits over the existing double deck trains are claimed. However, these claims by TfNSW are shown to be more conditional than publicly acknowledged; leading to the identification of key areas where more consideration and/or more information is needed.

The publicised raw capacity numbers, 30 single deck trains per hour carrying 1,300 passengers each (39,000 passengers per hour), compared with the 20 double deck trains per hour carrying 1,200 passengers each (24,000 passengers per hour) currently being achieved, substantiate the broad claim of more passengers per train and per line. Shorter dwell times (particularly at busy stations), more frequent services and less delays as loadings increase will clearly achieve shorter and more consistent travel times for single deck trains, even though the proportional saving falls as station spacing increases.

Digging deeper on capacity

The above capacity comparison depends on a greater acceptance of standing, and standing at a greater density, being assumed. Using seat estimates of 500 for single deck trains and 900 for double deck trains, the standing numbers would be 800 at 4/m2 (62%) for single deck and 300 at about 1.5/m2 (25%) for double deck at maximum capacity. For passenger standing acceptances below 62% the effective line capacity advantage of single deck trains would reduce towards a break-even point where the line capacities are equal and below which double deck trains have an effective line capacity advantage. Note that the 500 seat estimate for single deck trains is consistent with the modified Melbourne X'trapolis train seating adjusted for a longer train with less crew space and more doorways.

The break-even point can be derived from the double deck line capacity and single deck train design features. Running 30 single deck trains per hour to carry 24,000 passengers per hour equates to 800 per train with 300 (38%) standing at a density of 1.5/m2. This fairly modest increase in standing from 25% to 38% of passengers needed to achieve equal capacity without any increase in standing density, coupled with a travel time advantage and some specific benefits from using single deck trains for the future cross-harbour and CBD link as discussed below, could well have been enough for TfNSW to propose the use of single deck trains on this link and on the NWRL. However, this does not appear to have considered the potential line capacity that can be achieved by improving the operation of double deck trains.

Double deck potential capacity

A combination of modern signalling, wide platforms, good approach alignments and better dwell time management can all contribute to greater line capacity. These conditions are clearly not being met at busy Sydney CBD stations, such as Town Hall; with train operation there restricted by legacy signalling, narrow platforms, curved alignments and considerable contra flows (AM northbound boarders and PM southbound alighters), but are undoubtedly being incorporated into the NWRL.

Modern signalling allows trains to make a faster close up to an occupied platform, and to then achieve a shorter replacement time interval to enter that platform. The former reduces delays at busy times, while the latter also contributes to reducing the headway between trains. Good alignments, allowing maximum advantage to be taken of acceleration and deceleration, can also help with reducing headway. Wide platforms allow first alighters to move well clear of the train, reducing delays to those behind and to waiting boarders, with the consequent dwell time savings also reducing headway.

Under the above circumstances, an overall headway reduction of 30 seconds looks achievable, leading to an aspiration of 24 double deck trains per hour. A modest increase in the capacity of each train, from 1,200 to 1,300, should also be possible within this headway except possibly at stations where there are very substantial contra flows. The revised capacity numbers would then be 900 seats and 400 standing (31% at 2/m2) per double deck train and 31,200 passengers per hour. Note that it would be theoretically possible to double this standing density in an attempt to chase the single deck line capacity, however extended dwell times at busy stations would lead to little capacity improvement being achieved in practice and travel times would deteriorate. Additionally, the consequent high 47% standing level for double deck trains would seemingly defeat the purpose of their use.

It is also possible that the traditionally accepted 20 minute tolerance of standing may only apply to 2 standing passengers per m2 as this provides room to move for others to board or exit, and the ability to employ time usefully by reading or operating hand held devices. Accordingly, assuming 4 standing passengers per m2 (for either single deck or double trains) could lead to optimistic estimates of standing numbers being made because this higher density may be tolerable only for a much shorter period.

The revised break-even

30 single deck trains carrying 31,200 passengers per hour (the above aspiration for double deck trains) equates to 1,040 per train with 540 (52%) standing at a density of 2.7/m2. This more substantial increase in standing from 31% to 52% of passengers to achieve equal line capacity, coupled with an increase in standing density, presents a larger hurdle to be offset against (the somewhat reduced with better signalling) travel time advantages for the NWRL.

There remain, however, other advantages from using single deck trains on the proposed new cross-harbour and CBD link. The ability to handle steeper grades will mean stations can be built at less depth and alignments will be easier to achieve. The link can also be designed with stations that attract much of the substantial contra flow traffic in the CBD, for which single deck trains are better suited, leaving the existing double deck Western/North Shore route to serve more of a longer distance commute function that would then make the remaining passenger flows at the busy CBD stations predominately unidirectional. This move, along with the other headway reducing measures covered previously and possibly some station modifications, may be enough to enable 24 double deck trains per hour to be run on this route in the longer term.

The revised break-even evaluation above infers a greater divergence between train type requirements for the NWRL and the new cross-harbour and CBD link; making it more difficult to adopt a common design for both projects (some observers have proposed a compromise design). Network opportunities from such a divergence are discussed next.

Network opportunities

The longer term ability to handle 24 trains per hour on the double deck Western/North Shore route means that there would be enough pathways to run services to the NWRL, upper North Shore/Central Coast AND to restore the outer Northern Line via Chatswood at all times, provided a single deck all stations to Gordon service serving the Waverton and Wollstonecraft route is introduced via the new cross-harbour and CBD link. Attracting most of the existing shorter north side rail commutes in this way, and creating opportunities for new ones, to free up more of the double deck route for longer distance commutes would also be more consistent with the principle of using Tier 1 single deck trains to serve busy inner areas. A shuttle service between Chatswood and Parramatta, or other locations, would also be possible, provided new underground single deck platforms are provided at Chatswood to preserve the existing turn back facilities.


A comparison between modern single deck trains and Sydney's present lumbering double deck train operations appears to have influenced the decision in favour of single deck trains on the NWRL and a future cross-harbour and CBD link to which an operational nexus has been assumed but not explained.

However, a more realistic comparison using a higher performance target for future double deck train operations suggests a greater divergence between train requirements for these two projects, with double deck more suitable for the NWRL and single deck more suitable for the cross-harbour and CBD link, such that breaking the operating nexus between the two projects also needs to be addressed. The network benefits from breaking the nexus and introducing double deck operation on the NWRL in the longer term appear to be worthy of further consideration.

August 2013