In laying out the justification for the second stage of the $20 billion-plus rail project, a truncated version of its business case reveals that a suspended carriageway or a viaduct would have cost at least $400 million more than tunnelling under the harbour.

The two options for running a new metro line on the bridge also had a "number of constraints", such as posing a barrier to ships passing beneath or detracting from one of Australia's most famous landmarks.

"These options would also have broader network impacts during construction and operation (particularly in terms of access to the Sydney CBD for other transport modes). Consequently, these options were not progressed further," it said.

Another option to run the metro line along lanes seven and eight on the bridge would have required connections at either end of the bridge to tunnels for the metro line.

"Unlike the tunnel option [beneath the harbour], use of the Sydney Harbour Bridge would require the use of existing suburban rail stations and platforms at North Sydney and Wynyard," the business case summary said.

"The use of existing infrastructure for the project would largely result in replication of the existing T1 North Shore Line and would not provide additional rail services to new areas."

The government eventually opted for the second stage of the metro line to run under the harbour from Chatswood to Sydney's CBD, and on to Sydenham and Bankstown.

It has put a price tag of between $10.5 billion and $11.5 billion on the second stage, while the cost of the first section from Sydney's north west to Chatswood is $8.3 billion. The latter is due to open in 2019.

The business case summary also shows that the government considered a conversion of the existing Bankstown Line to allow it to carry metro trains to be one of the "less complex options" when it weighed up the route for Australia's largest rail project.

The Bankstown Line needed less infrastructure work to convert to a metro service when compared with other lines such as the South or Illawarra. The latter would have required extra tunnels and tracks, and "significant enabling works such as alternative freight routes", it said.

The conversion of 13.5 kilometres of track between Bankstown and Sydenham will be one of the most disruptive parts of the project. Tens of thousands of commuters will be forced to catch buses for more than six months during construction.

Critics have questioned the rationale for spending hundreds of millions of dollars on converting a line that is under-used by commuters.

They have also criticised plans for an extra 36,000 dwellings to be built along the Bankstown-Sydenham corridor over the next two decades.

But the business case summary, released on Tuesday, said converting the Bankstown Line between Sydenham and Bankstown would "improve network reliability by reducing the number of rail lines sharing the same existing tracks".

It would also "unlock capacity" at Central Station – the city's busiest - and "significantly reduce platform and train crowding".

"The T3 Bankstown Line does not share operations with other lines or rail freight. It would therefore be less complex to convert and segregate from the existing rail network when compared with other lines."

Labor transport spokeswoman Jodi McKay accused the Baird government of drip-feeding information about the multibillion-dollar project and called on it to release the full business case "which they have been sitting on for six months".

""It is clear the government has failed to learn the lesson from WestConnex, which has operated in secrecy," she said.

"It needs to come clean on how much fare revenue it will get; how it will be paid for, including how much will be raised by high-rise development above stations; and the level of disruption it will cause to residents along the Bankstown Line."