Commuter rail versus light rail, some costs examined

NZ in Tranzit goes riding the rails with an eye for the costs

Creating commuter rail, as part of the earthquake recovery in Christchurch, has been suggested in several recent newspaper articles and submissions to local hearings. It has been seen as a better alternative to light rail proposed down congested Riccarton Road. 

NZ in Tranzit believes the existing lines in themselves do not offer sufficient advantage for commuter rail and has suggested creating a proper network by adding links to new residential areas and other key passenger traffic generators. 

This posting is about trying to get an approximate compass fix on the cost of building new rail corridors around Christchurch.

While there is no way of knowing exact costs it is good to get a close approximation if only to discover all the secret snags or other realities that can upend a good idea.  I get sick of the absurd claims that various cities are the same size as Christchurch or talk of tram-trains that completely ignore the heavy freight factor in Christchurch (up to 16 full & empty coal trains a day) missing on these other lines, such as Mulhouse in France.

The light rail proposal being put forward for Christchurch at about $55 per kilometre appears fairly similar to the set up cost per kilometre of the light rail currently being built in Gold Coast City. That is after the $A180 million needed to purchase about 173 roadside properties and part of a further 111 properties  to a build a mainly separate lane structure for the trams is deducted from the total cost of stage one,  $A980 million dollars for 13 kilometres.  Much of the success of light rail (as with any transit mode!) is based on clear run and having totally segregated corridors or lanes (80% of of all French light rail is "off road, about the same proportion of Gold Coast light rail will be) -  the lower cost in Christchurch will not deliver much punch if running merely on streets in congested traffic.

Conventional rail appears to be considerably cheaper, not least I imagine because it does not have to dig up and replace, rebuild or strengthen multiple underground service conduits every few metres along the way, and can use the spread weight of shingle embankments to absorb weight and vibration.

I have identified five conventional or "heavy" rail projects in New Zealand and Australia that have been completed in the last five years that may offer some perspective on comparative costs.

These are  some broad figures deduced from (1) Perth's new commuter line to Mandurah; and slightly more precise figures gleaned from various sources regarding (2) the upgrading of the Wellington region commuter rail line between Pukerua Bay and Paekakariki; (3) electrification and double tracking for 13 km north Paekakariki to extend commuter rail to Waikanae. In greater Auckland recent projects include the reopening of (4) a commuter rail service to Onehunga and (5) the building of a short, commuter rail only, spur line from the main line to Manakau City.

All projects include site specific factors but I believe some general parameters of cost can be established for amateur trainspotters.

(1) The Mandurah line which opened in 2007 is a 70 km long suburban becomes regional rail corridor heading south of Perth which cost $1.6 billion to build, albeit this cost is particularly tilted by having 700 metres built underground in central Perth.

However an article on the well researched Melbourne Passenger Transport Users (PTUA) website comparing the costs of motorways to rail gives us an approximate fix for the basic line costs (albeit with overhead wiring).

We cannot rely on Melbourne experience for accurate costings of rail lines, since the last major urban rail extension completed in Melbourne was the Glen Waverley line in 1930. However, Perth is currently undergoing a renaissance in urban rail, and we can use their costings as a benchmark.... Perth's new southern railway to Mandurah was built for $12 million per kilometre, including the cost of freeway realignment and tunnels under Perth CBD. Excluding the latter, the cost of earthworks, track, overhead, stations and road overpasses for the 70km surface railway was $422 million, or just $6 million per kilometre.

(2) Wellington region's Kapiti Coast line snakes through several tunnels north of Pukerua Bay. In 2007 Finance Minister Cullen announced "About $80 million will be spent adding a second track to parts of the 3.4-kilometre section of the main trunk line between Pukerua Bay and Paekakariki".  As far as can be deduced this included day-lighting one tunnel (replacing it by a deep cutting) and dropping the floor level of three others, work carried out in 2010, so this doesn't give much of a fix except to say you get a lot of heavy engineering extras for around $NZ20 million per kilometre heavy rail as compared to light rail.

(3) Also on the Kapiti/North Island Main Trunk Line has been the extension of electric commuter rail north of Paekakariki to Waikanae. Ths involved double tracking an existing rail corridor and adding overhead wiring and rebuilding two stations, Paraparaumu and Waikanae (a station at Raumati originally mooted was never built). The distance was 13 kilometres and the cost was $92 million for the line and $6 million each for the two stations upgraded (including underground access etc). This is about $7 million per kilometre. Interestingly, I heard this described for some years as bringing regular commuter rail to Waikanae a fast growing area of 38,000 people - only recently working through census maps did I discover this population figure is for the whole district called Waikanae and includes areas that were already served by the existing line. Over a $100 million it appears was spent bringing new electrified commuter rail connection to less than 15,000 people, a residential population growing fast - but so too are many areas elsewhere around NZ including Christchurch.

Another curious fact is Wikipedia Kapiti Rail Line  states; "The project involved 50 workers and 20 machines installing 600 traction poles in eight or nine metre deep holes, and laying 30 km of rail and 30,000 sleepers. This suggests the existing line was also replaced. I am unable to trace this quote nor find any other reference to 30 km of track,  but if this is so costs would be closer to $3 million per kilometre.

(4) Upgrading the 3.6 kilometre rail line from Penrose to Onehunga to take commuter rail for the first time in 38 years, ostensibly was done at a cost of $13.6 million. This consisted of $10 million for general line upgrade by KiwiRail and $3.6 million by Auckland Regional Transport Authority on new stations at Penrose, Te Papapa and Onehunga. However I notice in research this does not seem to include an earlier expenditure in 2007 by the then Auckland Regional Council of $8 million. This was to purchase two hardware warehouses, demolished to make possible a site for a future Onehunga railway station. This line doesn't appear to have any major additional infrastructure and comes in at $6 million per kilometre upgrade cost (not known whether the tracks themselves were replaced but it sounds very like it)

(5) The latest addition to New Zealand's rail system - the first new rail corridor in Auckland for 80 years - is only 2 km long but expected to be extremely busy. This is line linking the centre of Manakau City [as it was until amalgamation into the Auckland "supercity"] to the Auckland commuter rail network and to central Auckland through a junction point off the North Island Main Trunk line at Puhinui. This project costs $98 million  but includes a 7 metre deep  300 metre long trench bringing the train in under roading in the central Manakau business area. A campus of the Manakau Institute of Technology will be built immediately above the line and station, starting with 1500 students in 2012, but planned to grow to 25,000 by 2040.

In Christchurch it would not be realistic to build a commuter rail network - adding dozens of trains a day to the already busy freight and coal train corridor - without creating some form of grade separation between Middleton yards and the existing overhead bridge at Durham Street. Traffic around rail crossings at Lincoln Road and Whiteleigh Avenue is already hugely congested without constant commuter train interruptions! Increasing mainline through tracks to three and/or trenching the rail lines is one option. Arguably Christchurch already has as strong a case for grade separation in this area, as strong a case I suspect as did Manakau,  as did New Lynn ($160 million for 800 metres trenched, albeit with a bus/rail station infrastructure included).

The figures and calculations here, for these five projects, can only be a rough and ready indication. Every over-bridge of the rail in suburban areas such as at Waterloo Road or at Buchanans Road on the Western Rail Corridor  NZ in Tranzit proposes will probably add $15-20 million and ideally at least 6-8 should be built as well as a northern motorway underpass, and of course buying land will add a few million dollars more as well to project costs.

Even so it is fairly clear from the rough scan of figures above that a basic new rail corridor - just track with level crossings - could probably be built at generous estimate for less than $10 million per kilometre - about one fifth of the cost per kilometre of the proposed light rail along Riccarton Road. Adding in attractive station designs, overbridges etc might lift this to $15-20 per kilometre - or about $260 million for the city links proposed (Redwood-Islington, 10 km; Northwood (Styx Centre) to Prestions 3.5 km). I have not added the suggested eventual Rangiora-Woodend-Pegasus-Kaiapoi loop as per this map but that is virtually all open country and no major river crossings.

Upgrading or adding additional mainline quality track on to the Hornby - Ferrymead section (suggested terminus at National Rail Museum) could add an unknown extra amount, but would enhance all rail operations and options. This might include about 2 km trenched or cut and cover section of the rail corridor, from Durham Street to Middleton - or an overhead rail viaduct - which could double or even triple this total cost, but may need to be considered anyway.  Or alternately several new road over-bridges such as at Lincoln Road ( perhaps half and half, the rail partially trenched, the road partially raised to get less ugly and divisive bridging than the current Colombo and Durham Street over-bridges).

It is hard not to think these huge amounts would be more wisely spent on upgrading and extending our existing rail infrastructure to enhance freight movements, road traffic movements and not least create a commuter rail network for Christchurch, rather than a far far more expensive per kilometre light rail system, delivering very little in comparable economic, social and environmental benefits.