How deep are Melbourne’s City Loop tunnels?

Melbourne’s City Loop – when racing to catch a train from the lowest level of Parliament station, it feels like a journey to the centre of the earth. But how deep are the stations in reality?

Climbing out of Parliament station

The City Loop consists of four tunnels over two levels: platforms 1 and 2 on the upper deck serve the lines through Clifton Hill and Caulfield, while platforms 3 and 4 on the lower deck serve the lines that pass through Burnley and North Melbourne.

The process

Unfortunately detailed diagrams of each City Loop station aren’t accessible online, which rules out the obvious way to read the depth measurement. Firing up the GPS on your phone could be another way, but getting a fix on the satellites from underground is impossible, and even if you could – the altitude accuracy of a consumer GPS device is about +/- 15 meters!

I ended up turning to another set of diagrams – the railway issued ‘Grades and Curves’ book that shows the general alignment and elevation for each of the City Loop tunnels. (the full set of diagrams can be found at Vicsig)

City Loop grades and curves diagram

The diagrams in the ‘Grades and Curves’ book are used to assist train drivers in learning the routes they will be driving along – trains don’t accelerate and stop like a car, so knowing about sharp curves, long inclines and steep descents is essential to keep the train under control.

Unfortunately the Grades and Curves diagrams only include elevation in metres above the Low Water Mark – a datum set as the level reached by seawater at low tide – and not in metres below the surface!

To determine the ground level elevation of each station, I turned to the survey marks database operated by Land Victoria. It gives the elevation of thousands of survey points around Australia, all given in relation of the Australian Height Datum (AHD) – a height set to the mean sea level for 1966-1968 calculated at thirty tide gauges around the coast of the Australian continent.

With the depth of each platform given by the Grades and Curves diagrams, the ground level elevation from survey marks, and the the rest is just simple arithmetic.

Flagstaff Station

Escalators at Flagstaff - three of the four banks only have 2 escalators and a set of stairs

For Flagstaff station the Grades and Curves diagram states the upper level is at +8 metres LWM, and the lower level is at -1 metres LWM. William Street is at the crest of the hill, so survey mark ‘MELBOURNE NORTH PM 50 MURLA’ (#308300500) on nearby Little Lonsdale Street can be used, located at a height of 28.446 metres above AHD.

Datum Calculated
Surface level +28.446 metres AHD
Rail level, upper platforms +8 metres LWM
Depth, upper platforms 20.5 metres
Separation between platforms 9 metres
Rail level, lower platforms -1 metres LWM
Depth, lower platforms 29.5 metres

Melbourne Central Station

Escalators from the Melbourne Central concourse down to platform 3/4

For Melbourne Central station the Grades and Curves diagram states the upper level is at +2 metres LWM, and the lower level is at -7 metres LWM. LaTrobe Street is a complicating factor – the Swanston Street end being higher than than that at Elizabeth Street – so I looked at two survey marks – ‘MMB 380′ (#450003801) on the Elizabeth Street corner is at 13.450 metres above AHD, while mark ‘MMB 388′ (#450003881) is located at a height of 22.370 metres above AHD.

Datum Calculated
Surface level +13.450 / +22.370 metres AHD
Rail level, upper platforms +2 metres LWM
Depth, upper platforms 11.5 / 20.5 metres
Separation between platforms 9 metres
Rail level, lower platforms -7 metres LWM
Depth, lower platforms 20.5 / 29.5 metres

Parliament Station

Collins Street concourse at Parliament station platform 1/2

For Parliament station the Grades and Curves diagram states the upper level is at +7 metres LWM, and the lower level is at -3 metres LWM. Spring Street is relativity flat atop the station, so I used survey mark ‘MELBOURNE NORTH PM 16 MURLA’ (#308300160) on Lonsdale Street, located at a height of 36.746 metres above AHD.

Datum Calculated
Surface level +36.746 metres AHD
Rail level, upper platforms +7 metres LWM
Depth, upper platforms 29.5 metres
Separation between platforms 10 metres
Rail level, lower platforms -3 metres LWM
Depth, lower platforms 39.5 metres

So what have I discovered?

  • Parliament station is the deepest below ground (39.5 metres) on the City Loop
  • Melbourne Central is the furthest station below sea level, but is the closest to the surface (between 20.5 and 29.5 metres) due to Elizabeth Street being in a valley
  • Flagstaff station is the highest station above sea level, but is the second deepest (29.5 metres) due to the location atop a hill

Checking my answers

After doing the maths myself, I remembered that the Metropolitan Transport Authority put together a booklet of City Loop facts and figures in the mid-1980s.

It quotes the following depths:

  • Parliament station: 40 metres
  • Flagstaff station: 32 metres
  • Melbourne Central station: 22 / 29 metres

I was out by 3 metres for the depot of Flagstaff station, but for the other locations, my roundabout way of determining the City Loop depth was reasonably accurate!

An alternate method

Another method to measure the depth of an underground station is to measure the height of an escalator step, count the total number of steps between platform level and the surface, and multiply them together to get the total height.

You can leave a response, or trackback from your own site.

9 Responses to “How deep are Melbourne’s City Loop tunnels?”

  1. mich says:

    The catch with your calculation, would seem to be that the AHD is not the same as the ( rather unusual ) “low water mark” on your rail diagram.

    • Marcus Wong says:

      AHD is what surveyors and builders use, while “low water mark” is a more colloquial term, and would presumably be lower than mean sea level indicated by the AHD.

      When applying my calculations I assumed that the railway use of “low water mark” was just a dumbing down of AHD for the purposes of simplicity – but it looks like I might have been wrong.

      This document goes into the relationship between various datums used in Victoria:

      http://admin.surveyorsboard.vic.gov.au/uploads/10/docs/Part%202%20Survey%20Procedures/SPH_S2_Sect8.pdf

      State Datum
      Mean Low Water Spring Tides was determined at Williamstown in 1884. This was related to the “broad arrow” cut in the sill (now covered) of the Williamstown Time Ball Tower. For 87 years, this mark and its value, 7.81 feet, was recognised as the basic bench mark defining State Datum.

      Railway Datum
      This was carried throughout the State by successive extensions of the railway system over a period of more than 100 years. Much of this took place prior to 1884 when the Railway Department adopted State Datum and there is no knowledge as to whether old records were amended.

      The above explains where the railway elevations are derived from, with the same document also indicating that the “Mean Lower Low Water” (MLLW) at Williamstown is -0.38 metres AHD – so that is how much my calculations were off by.

      This surveying thing sure is complicated!

  2. Tom the first and best says:

    It is interesting that the ramps, for the Caulfield loop, on the diagrams are marked as Caulfield and Sandringham not Dandenong and Frankston (as they function). The names mush have survived the cutting of the Caulfield flyover at Richmond (and Accompanying re-signalling to Caulfield). It was an unfortunate budget cut under the Bolte/Hamer Government that we are paying for today.

  3. Andrew says:

    Human instinct works well, as I have always judged how deep each station was by how long it took to get down to he platforms and my order of deepness was spot on. Nevertheless, well done you to come up with the stats.

  4. mich says:

    You’d think, if they were worried about the tunnels, they’d be more interested in the high tide level, that is when you are going to get back-flooding through the drains. But I don’t know if the tides in Port Phillip Bay are as big as other places, if your discrepancy there is only 38 cm, then perhaps not.

Leave a Reply