Safe operation of the railway starts at the design stage.
Involving operators in the project from the outset is essential to designing a railway that is safe, says KCRC operations and maintenance general manager Wilfred Lau. 'We are learning from the bitter experiences of other railways in the world,' he says. 'A lot did not have any operators on board at the design stage.'
Engineers pooling their experience always try to give due consideration to safety, he believes, but often things do not work perfectly.
'We are taking a risk based approach,' explains Lau. 'We want to identify all possible hazards at an early stage and incorporate all the mitigating measures into the design.'
Lau's team includes operations and safety engineers as well as systems safety manager Bob Lupton. The team explains to the designers exactly what it wants. 'We go through everything, and come up with the user requirements,' explains Lau. 'This covers the system as a whole - train frequencies, where we want the station entrances, the communications network and so on.' Ease of maintenance is another factor.
'The whole purpose is to ensure that when someone gets on the train at Tuen Mun and gets off at Yen Chow Street, their exposure to risk is as small as it can be.' It can never be zero in an operating railway, Lau adds.
East Rail is one of the safest railways in the world, with 0.5 incidents per million passengers. 'We are aiming to be better than that,' adds Lupton.
This is probably the first time a full risk based approach has included operators at an early stage, says Lupton. Some systems such as Docklands Light Railway started the process a little later, he says, while others have adopted a risk based approach but without involving operators.
The process involves identifying the hazards, quantifying and assessing them, taking action to reduce the risks and reviewing the decisions. 'One thing we hammer home to everybody is that it is all very nice doing this theoretical work, but if we don't take practical action in our design it is all wasted,' emphasises Lupton.
The system as a whole was modelled prior to detailed design, and is examined in modules as the design develops, before the whole is integrated as the system is built. Risks are reduced to 'ALARP' - as low as reasonably practicable - bearing in mind financially viability.
Deaths on the railway, excluding suicides, are fortunately rare. But minor injuries - cut hands and bruises are more common, and occasionally there are major injuries, say when someone falls down an escalator. 'We have to consider all of those,' says Lupton. The concept of equivalent fatality is used - one equivalent fatality means one death, 10 major injuries or 200 minor ones.
Some values are clear cut: a risk of greater than about 10-5 per year - equivalent to five fatalities - would be totally unacceptable to KCRC, says Lupton. 'We would spend money however much it cost if this was the case.'
At the opposite extreme, the railway would be the safest in the world if the figure was 0.05 equivalent passenger fatalities a year - or one every 20 years.
But in between these lies the ALARP region, where money is spent depending on the value of fatalities that could be prevented. The value of a fatality is put at HK$15M (£1.2M). 'If we knew we could save one person a year by spending up to HK$15M then we would do it.'