Unfair - cars deserve the blame. The buses were delayed by the cars. Before buses were removed, did you ever see the George St entertainment area on Friday evenings at about 6pm? Inbound buses could barely move even though they were going in the opposite direction to the official peak and were less numerous than in peak. We should have removed the cars first.
In the Environmental Impact Statement ("EIS") for the Sydney CBD & South-East Light Rail project ("CSELR"), Transport for New South Wales ("T4NSW") claimed the ultimate capacity to be 30 trams × 300 pax = 9000 pax each way per hour. That was a serious exaggeration. A more useful figure would have been 24 48-metre trams × 250 pax = 6000 pax each way per hour. The initial capacity in 2019 was be 20 trams × 250 pax per hour = 5000 pax each way per hour of which only 1600 pax had seats however subsequent announcements foreshadowed a rise in capacity. The Alstom website shows that typical configurations provide seats for less than 30% of the passenger capacity. The 466 passengers claimed for 67-metre trams should likewise be discounted to 400 with just 96 seated (the 24 hinged seats are assumed to be in the folded-up position, making room for 2 passengers to stand rather than for one to sit).
The claimed capacity for 30 trams per hour each way seems unachievable at peak hour, just when it would be needed. It would require the major intersections on the route (Bridge/George/Grosvenor, Elizabeth/Devonshire, Anzac/Cleveland) plus the Alison Road crossings to be interrupted up to 60 times per hour for a tram to cross. Co-ordinating the inbound and outbound trams reliably and thereby holding the disruptions down to 30 times per hour would be impossible. See remarks below about interference with cross traffic.
Whether that capacity would be enough for the load in the morning peak is academic. The actual capacity will be much less - at most 6000 before overcrowding becomes unacceptable. The inbound trams will be overloaded as they approach the Chalmers St stop. Worse, many passengers entering the CBD on Broadway buses would find a tram ride up George St preferable to a slower and longer bus trip along Elizabeth St. These passengers could be expected to transfer to tram at Rawson Place however there probably won't be enough room for them.
The situation might be even worse than that. There seem to be grounds for suspecting the EIS figures were massaged to downplay the overcrowding. For instance, the EIS statistics were presented in whole hours, possibly masking a greater peak between perhaps 0740 and 0840. In any case, there will inevitably be fluctuations in trams' loadings depending on how many busloads they accept at Randwick or Kingsford, making crowding in some of the trams even worse than averages would suggest.
Worse again, the EIS showed that the planned Kingsford leg will attract more passengers than the Randwick leg. This imbalance will exacerbate the overcrowding on half the services.
About 40 public submissions to the EIS vainly questioned CSELR's capacity. Peter Mills' submission looks at several aspects of capacity. The official story that capacity was adequate changed on 23 October 2014 when a ministerial announcement said "the biggest challenge for this project is meeting the high demand" and went on to foreshadow a substantial increase in capacity although by unspecified means. On 18 December 2014 another ministerial announcement lengthened the trams from 48 metres to 67 metres but was vague about what this meant for capacity.
The eventual agreement with Alstom was for a maximum of 15 services per hour each way along George St, one every four minutes. That means one tram each way every eight minutes on each of the Kingsford and Randwick legs. Assuming the claimed capacity of 466 passengers per tram, 15 x 466 = 6990 pax per hour. That capacity would be barely enough for the peak-hour figures shown in the 2012 diagram above. The demand encountered in 2019 will be more. Further, there should be an allowance of perhaps 20% for fluctuations in demand. So not all the 2019 passengers will fit within the available capacity. Presumably the rest will be accommodated in extra peak-hour express buses and/or squashed onto trams at more than 4 per square metre. The 15 trams per hour will have only 120 seats each (96 fixed and 24 flip), so only 15×120 = 1800 of the passengers can be seated.
Here is Table 29 from page 33 of Transport for NSW's capacity planning guidelines dated December 2013:
Just for comparison, the seated capacity of the replaced bus services (the EIS says 220 morning peak hour buses won't have to enter the CBD) would be at least 45×220 = 9900 seats.
Developments in the pipeline stand to increase the local population, and hence the demand for transport, substantially. For example, a proposed development of 750 home units on the former Inglis stables site has been approved. There will be other approvals.
Pursuant to a 2013 plan called Sydney City Centre Access Strategy, which you can download at http://www.transport.nsw.gov.au/sites/default/files/b2b/publications/sydney-city-centre-access-strategy-final-web.pdf, from 4th October 2015 all George St buses (including those that use the Anzac Bridge) were diverted. Many now use Elizabeth St to travel to Circular Quay; some go elsewhere. Elizabeth St is presumably now a slower trip than at present because Elizabeth St was already congested and would have become more congested with the extra traffic. These detours will continue through the construction and testing phases of the CBD and south-east light rail project - at least three years.
Once the CBD and south-east light rail service is operating in perhaps 2019, all of the former George St bus services will be altered to either terminate at Haymarket or else leave the city in other directions. All passengers wishing to ride further into the CBD beyond Haymarket will have to transfer to light rail.
However, the northbound light rail vehicles arriving at Haymarket will already be carrying passengers from the south-eastern suburbs and will probably be unpleasantly crowded in peak hour.
Please note that the Sydney City Centre Access Strategy is officially a quite separate plan from the CBD and south-east light rail and therefore no-one in authority will admit that it is a consequence of the light rail decision. The strategy has a rather Orwellian view of transport - for example, its December 2015 update says the disruptive 4 October changes were to "improve the way we use our city centre".
We don't know much because Transport for NSW hasn't released any analysis. But we do know that they are concerned - in a submission on the proposed Barangaroo casino T4NSW says "the changes to George Street in the Sydney CBD when light rail is built there will increase traffic movements on the road work located within and adjacent to the Barangaroo project". See http://www.smh.com.au/nsw/barangaroo-casino-parking-traffic-concerns-raised-by-transport-for-nsw-20150515-gh2cpo.html.
Broadway-bound passengers from around Wynyard will quickly fill available space in southbound George St trams because they will thus get to Rawson Place sooner than Elizabeth St buses can take them. Eastern suburbs passengers will therefore find boarding difficult around Town Hall and will be delayed waiting for a tram which can carry them.
Doubt has been expressed whether the A.M. peak service can handle the combined outbound load of SBHS, SGHS and UNSW passengers. And the principal of SBHS has expressed concern that an afternoon peak of returning USNW students might load trams to the point where some high-school students are unable to board.
A count made in Albion St on 28 August 2014 found over 3600 passengers on 891 (UNSW) and 610 (SBHS, SGHS) riding between 8 a.m. and 9 a.m. Even without any passengers for other destinations, not many of the students can expect seats on trams - there will be only 1770 seats per hour.
Especially in the morning peak, there will be situations where adults and schoolchildren are standing together in tightly-packed trams. Parents might well be reluctant to let their children ride in these conditions.
T4NSW figures seem to rely on four standing passengers per square metre of floor space. If you can't visualise that, think of a fully-loaded lift in which people stand at about that density for a few seconds per trip. What if most of those people are students with backpacks?
How many buses do the trams replace? The June 2014 video says 220 - see http://www.youtube.com/watch?v=pUQoV8DPkNI. If that means 220 per morning peak, which routes are they? The EIS nominates several routes which are to remain bus but points out that the list might change. There aren't 220 buses in the other south-eastern routes.
Table 2-1 (EIS Technical Paper 1, page 39) lists services under study, including express services.
EIS Section 4.2.2 provides an overview of possible bus service changes. Some bus routes are to continue to operate into the CBD. However, the Cleveland/Crown district will lose most of its bus services yet be well away from light rail.
Sydney's Light Rail Future (dated December 2012, page 18) provides further information but may be out-of-date. The figure of 220 buses shown above is close to the 227 shown on page 18 but that 227 includes 55 Harbour Bridge services!
About 1084 scheduled buses arriving from Broadway pass through Railway Square each weekday, 159 of them between 8 a.m. and 9 a.m. These figures do not include red metrobuses M10 and M30 (six of each per peak hour). Nor do they include L88 and L90 etc. buses and Hillbus services which join George St further north.
Sydney public transport is too slow and is not getting faster. Many rail services, including the Dulwich Hill light rail, could and should be run at higher speeds. Professor Hensher has published research showing that replacing M2 buses by feeders to Sydney Metro North-West will result in longer trip times to the CBD from most of the north-west.
CSELR speeds as envisaged by the EIS will also be unacceptably slow. For example, 15 minutes from Central to Circular Quay compares very unfavourably with heavy rail, which can do the whole city circle from Central to Circular Quay to Central with 5 intermediate stops in about 13 minutes. CSELR's slowness is probably due to insufficient priority over crossing traffic.
The popular trip from Central to UNSW is also likely to be slower in CSELR trams than in existing express buses.
It has been pointed out that, based on a number of European examples plus the Gold Coast Linq and noting the relatively low number of CSELR stops, that a trip of 8.5 km from Circular Quay to the end of either CSELR branch should take about 22-23 minutes and certainly no more than 25 minutes. T4NSW estimates have been considerably longer - about 35 minutes. This discrepancy can only be explained by lack of traffic light priority.
The former trams withdrawn in the 1950s got between Circular Quay and Central in 15 minutes with 12 intermediate stops (one at nearly every intersection). CSELR is projected to take 15 minutes over this section with only 5 intermediate stops. Again, the only reason can be lack of tram priority at traffic lights on cross streets.
Some trams will get lucky and encounter a series of green lights. Unfortunately, that doesn't mean they'll be much quicker. Unless the tram in front has been equally lucky, the fast tram might find itself running up against the earlier service. It will have to slow down, not only because overtaking is impossible but also because the tram stops can only handle one vehicle at a time.
Light rail vehicles typically accelerate and brake at about 1 metre per second per second. Any more is uncomfortable to standing passengers and is likely to result in passenger falls inside the vehicle. The Alstom website shows maximum braking and acceleration are both 1.2 metres per second per second for the Citadis X05.
This has particular significance for cross traffic. For example, the offset Bridge/George/Grosvenor intersection spans 55 metres of George St. A 67-metre light rail vehicle must therefore travel 55 + 67 = 122 metres to clear that intersection. This may well be from a standing start if the traffic signals were against the vehicle when it arrived. And southbound vehicles will have to stop immediately after clearing Grosvenor St. Let us optimistically assume the vehicle starts immediately it gets a green light, accelerates at full power and brakes at maximum, and that the traffic signals change to favour cross traffic immediately after the vehicle has cleared the intersection. Formulæ taught in high school show that about ten seconds is required for acceleration during which the vehicle travels 61 metres. Another 10 seconds is required to decelerate over a further 61 metres. So just over 20 seconds will be required for each vehicle to cross. In practice, each interruption to Bridge/Grosvenor traffic would be rather more than 20 seconds.
Grosvenor St is a major feeder to the Sydney Harbour Bridge and will increasingly carry Barangaroo traffic. Its importance to Harbour Bridge traffic was amply demonstrated on 5 January 2015 when an unplanned closure of the Bridge/George/Grosvenor intersection in peak hour resulted in buses and cars backed up across the Bridge to North Sydney, despite the light holiday traffic.
There are several other intersections where conflicts between light rail and other traffic are likely to be an issue at peak hour.
All of the stops, and some city blocks, are too short to contain more than one 67-metre light rail vehicle. Some city blocks are too short to contain even one light rail vehicle. Therefore the vehicles cannot travel in clusters and can only cross intersections one at a time. This reduces the number of LRVs that can be run each hour.
Infrastructure NSW prepared a State Infrastructure Strategy "First Things First" in 2012 which stated that using 60-metre trams was not considered feasible in Sydney's CBD as it would be too obstructive for retail loading/access points (page 98). What about 67-metre trams?
Bus services are readily diverted when large crowds, such as New Year's Eve, preclude the ordinary route. Parades (e.g. Anzac Day) and breakdowns are also handled by diverting buses. The light rail service must cope with these contingencies which it can only do by turning vehicles back before they reach Circular Quay. It might have to do the same in peak hour due to heavy cross traffic at Grosvenor and Bridge.
An incident on 25 February 2016, in which police closed Eddy Avenue for three hours to all road traffic and also stopped the light rail service which crosses Eddy Avenue, highlights the vulnerability of rail services to blockages. Buses were diverted away from Eddy Avenue during this incident but were otherwise able to operate. CSELR light rail services would have been cut at Central for the whole duration of the incident.
Gold Coast light rail services are prone to disruption too - see, for instance, http://www.brisbanetimes.com.au/queensland/gold-coast-trams-suspended-after-collision-20161009-gry5sm.html. And on 15 January 2017, a tram derailed after striking a fire truck, blocking both tracks. If the derailed tram had travelled one metre further, it would have struck a pole which supported overhead power lines.
A strategy to turn George St into "Sydney's major boulevard and a world-class destination for walking, shopping and dining" has been published in draft as George St 2020. It is unrealistic - the authors have overlooked such boring details as maintaining essential access for daily shop deliveries, occasional heavy deliveries and emergency vehicles.
It is suspected that T4NSW has found by modelling that giving the trams the priority they need at junctions like Cleveland St, Anzac Pde and Lang Rd would result in serious detriment to traffic conditions over a wide area. However, no modelling has been released.
Sydney has a long history of parkland being filched for other purposes. This can happen gradually over many years. While sport might seem a valid reason for appropriating parkland, consider the case of Wentworth Park (between Pyrmont-Ultimo and Glebe), much of which is now a greyhound track and associated grandstands.
Further west, the eastern corner of Parramatta Park was long ago handed to the Parramatta Leagues Club. Fans celebrating the club's premiership win in 1981 started a fire which destroyed the old Cumberland Oval grandstand nearby. The club then wanted to replace the oval by a 40000-seat stadium. Both the local council and the park trustees approved however the project was limited by the Land and Environment Court. Twin grandstands were erected and became known as Pirtek Stadium. They were demolished early in 2017. The club is expanding Pirtek to surround what was once a parkland oval. The adjacent Parramatta Swimming Centre (opened 1959) and associated car parks also stand on former parkland. At the south corner, so does Parramatta RSL Club. The swimming centre site will be absorbed by the Pirtek expansion.
Note that in 2001 the proposed Epping-Chatswood rail bridge across the Lane Cove River was to be approved on the condition that there be a land swap for lost parkland, even though the affected land already had a busy road bridge across at the same point. (The Parramatta Rail Link was approved in February 2002, passing under the river in a tunnel and hence not requiring parkland.)
Returning to CSELR, the following parks are affected by the project:
High Cross Park was selected for an above-ground power substation incorporating "driver amenities" with a footprint of 80 square metres and which would probably be about 3 metres tall with noisy exhaust fans. However, the 17 September 2015 announcement said that the "transformer would now be underground" but didn't mention the $3 million undergrounding cost to be borne by Randwick ratepayers. The amenities will move to a location in High St near the terminus.
Including street trees in Devonshire St, Anzac Pde, Alison Rd and Wansey Rd etc, plus parks, the total loss is well over five hundred trees. Unfortunately, so many trees will be lost that anyone criticising CSELR on any ground risks being dismissed as a tree-hugging NIMBY.
About 100 residents were evicted by August 2014 and the building was demolished early in 2015.
People are not satisfied by "compensatory" parking to be provided remote from where they want to park.
Although it is possible to reduce wheel squeal on curves by treating the wheels and track, little can be done to reduce rolling rumble noise. This is likely to be an issue in Devonshire St and Wansey Road.
Neither of the relevant stops (Strachan St and the Kingsford terminus) is convenient for most of the existing shops. Strachan St is too far north. The terminus is too far south - crossing the Nine Ways intersection to get to it will be a nightmare for pedestrians.
Businesses are concerned about the loss of parking.
The trams use significant power. Indicative figures for the Gold Coast light rail system which uses 45-metre trams, are 1250 amps traction current plus 250 amps for air-conditioning etc, at 700 volts DC. Scaling this up for 67-metre trams suggests about 2000 amps. While full power is only drawn for short periods of acceleration, each substation has to be capable of supplying well over a megawatt for each tram in its zone.
None of the above makes wire-free operation a good idea in George St. For a discussion of the cons of wire-free operation, see page 261 of Transit Australia, September 2014. The system used by Alstom for wire-free trams relies on a segmented third rail which is only powered when there is a tram above it; relying on that may have the disadvantage of precluding trams by other manufacturers. Also note that the "ugliness" of overhead wiring can easily be exaggerated by a photographer with a telephoto lens; the wiring probably won't look as bad as depicted.
Interestingly, the Newcastle light rail service is also to be wire-free between stops. But it won't use the Alstom system because, according to the Minister as quoted in The Newcastle Herald, that system can't be used in Newcastle "because of the corrosive effects of the sea water". He didn't explain why Circular Quay water is less corrosive than Newcastle's.
Simply, make permanent the interim bus arrangements that commenced in October 2015. And don't build CSELR.
Better, all of the above problems could be averted by cancelling CSELR and instead extending the Eastern Suburbs railway underground from Bondi Junction to Kingsford. There would be a station servicing UNSW and the PoW hospitals. There could also be a Charing Cross station.
The question of banning cars from part of George St is quite separate. It could be done at any time. Or a toll cordon around the CBD could be implemented to deter cars from entering the area.
A 2011 Legislative Council inquiry found that NSW construction costs per distance for heavy rail were high compared with similar projects elsewhere in comparable economies.
According to SMH of 10 November 2014, estimated construction costs have blown out to $2.2 billion. However, important details such as the length of trams and the frequency of services were still being negotiated. Despite about forty EIS submissions criticising the project for lack of capacity, this inability to carry peak-hour loads acceptably was only admitted on 23 October 2014 when the Minister said "As I have said previously, the biggest challenge for this project is meeting the high demand from customers who are expected to choose light rail over buses, private vehicles and other transport options". So we cannot be sure we are hearing the full truth.
The risk will lie with the NSW government, not the winning tenderers who will have a guaranteed return.
This could have been implemented at any time. It is a separate issue from light rail.
Will tram windows be covered with perforated advertising film which becomes translucent in wet weather and obscures the view? See the problems set out in Wikipedia.
See Peter Mills' EIS submission and the table notes at the end of Peter Egan's submission.
It would be very interesting to see details of the calculation of 10000 jobs to be created by the light rail and also of the claimed $4 billion benefits.
The admitted costs do not include any allowance for disrupting Broadway and Anzac Bridge bus services nor for reducing passenger amenity in the eastern suburbs. If they did, the project would be much less attractive and might well be unwarranted.
The NSW Auditor-General reviewed the claimed business case and reported in November 2016 - see Recent Developments below.
The plan is to remove all buses from CSELR's service area, However, that would leave holes in the present services to some areas. For example, many of the buses which will be removed use Cleveland St to get from Randwick to the CBD. What happens to Cleveland St without buses?
Surprisingly, many people in the south-eastern suburbs still think that the light rail will supplement their bus services. It won't. As far as overcrowding will permit, bus services will be permanently withdrawn. And many people in the inner western suburbs don't realise that they will be expected to change between bus and tram at Rawson Place.
The authorities must be aware of this misunderstanding. We can only assume that they don't propose correcting it until the last minute if at all.
Originally, most George St shopkeepers seemed unconcerned about the growing disruption and its possible effects on their businesses. The Sydney Business Chamber asserted its unreserved support for the project. Yet the Chamber has also said "we ... need to take action to spread peak demand", presumably by facilitating increased flexitime which could reduce peak-hour demand for transport.
But many businesses, especially those which rely on passing trade, are suffering a downturn. This is resulting in some staff losing their jobs and in the closure of some business. The NSW Government is resisting attempts to extract adequate compensation from it, even as chronic construction delays make the pain worse.
Some shopkeepers have attended meetings organised to address the matter. The relief being offered will obviously be too little too late.
The march of CSELR will delay completion of the cycleway network - see http://www.smh.com.au/nsw/delays-threaten-sydney-cycleways-as-changes-to-castlereagh-street-considered-20141112-11l12e.html and http://www.smh.com.au/nsw/bike-lanes-out-as-george-street-light-rail-project-takes-off-20150528-ghbamo.
The serviced area outside the CBD almost coincides with the Randwick local government area, plus the two Sydney high schools. However, there are serious ramifications for the inner western suburbs, due to the effect on Broadway and Anzac Bridge bus services.
Relevant state seats in the east are Sydney (independent), Coogee (Liberal), Maroubra (Labor), Newtown (Green) and perhaps Heffron (Labor). In the western suburbs, the affected electorates will be Balmain (Green), Newtown (Green), Summer Hill (Labor) and Drummoyne (Liberal).
Some of the cost of the light rail is being paid by Sydney City Council. The Sydney local government area includes considerable areas in suburbs like Glebe and Erskineville and collects rates from those areas. However, the residents of these areas cannot be said to benefit from the project.
Many politicians think that investment in public transport is a good idea. However, CSELR isn't about getting passengers out of cars onto improved public transport services but rather shifting them from one mode of public transport to another at huge cost. Think of greenwash.
Under the former Goodling administration, innovative tendering was to be used to minimise construction disruption and to expedite the work. Tenderers for each piece of work were to stipulate not only their price but also the number of weeks they would complete the work in. Preference was to be given to tenderers who would work quickly. There was to be a financial incentive for keeping to the agreed schedule.
It is very clear that this didn't happen. The mess in George St speaks for itself. And trackside residents of Anzac Parade tell stories of hoardings going up, piles of sand and other materials lying abandoned for months, and then noisy work running overnight. Complaints about unnecessary noise are rebuffed by contractors who say they are allowed to work around the clock and the job was urgent.
Premier and former transport minister Gladys Berejiklian has not explained when she became aware of a construction cost blowout nor why she did not inform the public of the correct reason for the blowout. One wonders whether and when she ever realised she was building an expensive transport system that would be incapable of meeting its core requirement.
Despite recognition from many quarters that the light rail's capacity will not be adequate for peak hour, no admission has ever been made by Transport for NSW and/or the ministry that there is a problem and that solutions should be sought. It would be worthwhile seeking a second northern terminus for the line to supplement George St. Possibilities such as Bondi Junction, Kings Cross, St James, Green Square and/or Redfern/Darlington should be evaluated. Unfortunately, there is no indication that any such investigation is in progress.