Continued Occurrence of Train Breakdown Incidents

Mr Hri Kumar Nair asked the Minister for Transport how does the disruption rate of train services in Singapore compare to similar systems in other major cities, taking into account factors such as age, load and frequency of services.

Mdm Speaker, I fully understand and empathise with commuters' frustrations over the recent spate of rail disruptions in January at Kranji, Ang Mo Kio and Tanah Merah. I had called in SMRT's senior management to convey my grave concerns and find out the reasons behind the spate of disruptions and what SMRT was doing to ensure that this was not the start of a new trend of disruptions.

I was especially perturbed because I knew that there had actually been a lot of hard work put in by the team at SMRT as well as the SBST team, together with LTA, to improve train reliability over the last two years. And we had actually made good progress in so doing. I monitor two indicators very closely. Firstly, the number of delays that are more than five minutes; and secondly, the number of unplanned train withdrawals, because some of these train withdrawals may be shorter than five minutes, and yet if it occurred during peak hours, it can be very disruptive. That is also a very pertinent indicator that I ought to pay close attention to.

Page: 27

On the first indicator, last year, we reduced the number of train delays longer than five minutes by 20% across the whole MRT network, from 1.49 delays for every 100,00 km travelled in 2012, to 1.18 in 2013. The number of delays in 2013 was lower than that achieved back in 2010. What it means in terms of the number of trips that are made every day on the main lines – the North-South East-West Lines as well as the North East Line – is that it works out to just under one such delay for every 2,000 train trips that are operated on these two lines. Put in another way, 99.95%, or except one out of every 2,000 trips would not experience any delay, or a delay of less than five minutes.

On train withdrawals, the rates for the North-South and East-West Lines have also fallen, from 3.32 withdrawals for every 100,000 km travelled in 2012, to 2.25 withdrawals in 2013. This is about a 30% improvement. Again, what we have done is to have restored the train withdrawal rates close to the 2010 levels. For this year, the target for the North-South and East-West Lines is to bring this figure from 2.25 further down to about 1.8.

Likewise, we have also seen a more than 25% reduction in train withdrawal rates from 2012 to 2013 for both the North East and the Circle Lines. Indeed, the withdrawals for the Circle Line are at its lowest since it first opened in 2009.

It is not easy to compare service disruption rates across different countries and systems, in part because of differences in age, length of networks, loading frequency of services and so on, but more so because such data is generally not made publicly available by the operators, unlike our practice which is that every three months, we put this data out on the LTA website, so that Members of Parliament and interested members of the public can actually scrutinise the data comprehensively.

But I note that that for each of the past two years, Hong Kong's MTR – which has a very good record – its domestic rail lines have experienced about three to four major service disruptions of more than 30 minutes, which is similar to SMRT's North-South and East-West Lines. But, again, obviously the size of the network, the loading, the frequency of services are not quite the same.

The improvements in reliability we are seeing have come about because the operators have invested substantially more in maintenance and system upgrading, especially over the past two years. We estimate that SMRT has increased its annual repair and maintenance expenditure for its two rail lines by about 65% over the course of these two years, with $64.5 million spent in FY2013 compared to $38.3 million in FY2011. Even when we normalise the

Page: 28

repair and maintenance expenditure per train-kilometre for the North-South East-West Lines, so as to reflect the higher usage as well as the ridership that had taken place in these lines, the maintenance and repair expenditure had gone up by over a third in these two years.

It clearly reflects a considerable step-up in resources and efforts to intensify maintenance and fix reliability issues. Beyond this, the LTA-SMRT Joint Team is working on other measures to upgrade the trains and systems of the Compass Lines, which will involve major investments of about $500 million to overhaul the older trains by 2016 and upgrade the signalling system by 2018. Likewise, SBST has stepped up improvement works to rectify tunnel leakages and change out problematic components of the overhead power system in the North-East Line tunnels. It will be commencing a comprehensive mid-life upgrading programme for the North-East Line trains, which are just over 10 years old. The two operators have also ramped up their engineering capabilities and processes, and hired more engineers as well as expanded their maintenance teams.

On LTA's part, we are upgrading the rail infrastructure to improve reliability. The replacement of ageing timber sleepers with concrete sleepers on the North-South and East-West Lines has been under way since 2012, while the replacement of traction power cables on the Circle Line has been completed. These traction power cables were the cause of a series of disruptions back in 2012. Some of you may remember this. In 2014, the North-South East-West Lines' third rail system will undergo a full, system-wide change-out. We will also be improving the North-East Line's overhead power system with new corrosion-resistant materials, to minimise the recurrence of stress-corrosion cracking in its components that afflicted parts of the tunnel, particularly to the south, back in 2012 as well as the earlier part of 2013.

So, we have, overall, made some strides in the right direction, but there is certainly room for further improvement. Thus far, the joint teams that LTA has set up with SMRT and SBST have adopted a fine-and-fix approach to rectify existing issues with the rail systems. Following the completion of separate consultancy studies that we have activated on both of the operators, both operators will, going forward, adopt a new predict-and-prevent approach to their respective maintenance regimes. Under this new approach, operators will undertake active condition monitoring of rail systems and formulate maintenance processes to optimise the performance of the assets over their life cycles.

Page: 29

In our perspective, we understand that service disruptions do happen from time to time, especially as the rail network expands and ages, and as we run more and more train trips. But in the case of the January incidents, SMRT could have done much better in service recovery, especially to provide timely information to affected commuters. The LTA's Code of Practice requires operators to put out announcements at stations and through various media platforms to provide clear and timely status updates and to advise commuters on alternate travel options, such as the free boarding of public buses and shuttle services between affected stations.

As part of its investigations, LTA will highlight gaps in the SMRT's communication with affected commuters and work closely with both operators to share learning experiences and to further improve their incident management. The SMRT senior management team has already committed to review its contingency plans on a station-by-station basis and to refine its command-and-control structure in times of disruption so as to provide better ground support to its staff.

Madam, I have two supplementary questions for the Minister. First question: the disruption of 20 January 2014, which was at Ang Mio Kio station, had affected many of my residents because they were travelling to the city to work. I understand that the disruption was for more than 20 minutes and the residents said that there was no communication, they did not know what happened and it was very stuffy – no ventilation. I would like to ask the Minister what can be done because the train operator has been doing this all the time and still such things happened.

My second question is: it is good that the Minister called in the higher management of the train operators when this thing happened. I would like to ask the Minister whether the Minister would call in the higher management again to ask them to forgo the fare increment which has been recommended by PTC because the service is still not satisfactory.

As mentioned in my reply, certainly for the recent spate of disruptions, we do find shortcomings in how it was handled by SMRT in terms of the announcements not only to the affected commuters but also in making sure that people who are joining the network understand where the disruptions are, approximately how long it may take to recover, as well as to find and offer alternative modes of transportation.

Page: 30

On fare increment, it is best to separate these two issues. While I understand the emotions involved in it, we also have recently completed the Fare Review Mechanism Committee (FRMC) report. The FRMC report came up with a formulation that all parties ought to try to abide by and follow, which is to compute the fare adjustment on an annual basis, with room to roll over by taking into account the macro factors of wage increases, core CPI, as well as the energy index.

I thank the Minister for his response. Just two supplementary questions. One, in terms of communications, it is not the first time that we have had disruptions over the last three to four years. However, it seems that every time something happens, the transport operators are reacting to remedial situations that are happening when they could easily have projected them. Is there any reassurance from the Minister's meeting with higher management that they will take this into serious consideration? Two, the Minister mentioned the condition monitoring system that the companies are to introduce. Is there a timeline that we can see this put in place so that we can reduce further disruptions in the transport network?

I thank the Member for the supplementary questions. I meet up with the senior management of SMRT and SBST roughly on a quarterly basis to run through reliability issues, what they have managed to accomplish and what is still outstanding, to urge that they put in greater resources, for example, for re-sleepering, resignalling works and so on.

My feel after talking to the SMRT team is that they take it seriously. As the CEO himself told me, the team was actually very disappointed themselves with the series of incidents as well as with how staff on the ground had handled them. We agreed that SMRT should review their contingency plans on a station-by-station basis rather than to have a more generic plan that covers stretches of their rail system because you do need very specific drawer plans that you can activate at a time of contingency.

On condition monitoring, which I believe is the Member's second supplementary question, already some of that has been done. For example, there are a couple of trains that have been instrumented – these are operational trains so, therefore, they run across the entire network, that is, the North-South Line and the East-West Line – to, for example, detect whether there are sags in the third rail. Indeed, there was a recent incident where we had to require the train to, fortunately, make a bypass through the Bishan Depot because one of these trains picked up a minor sag in the third rail around the Ang Mo Kio-Bishan area and, because we were able to pick it up in time, the team decided

Page: 31

that they would want to rectify it as quickly as possible, and we were in a position to be able to do a bypass for the train. Within about 10 minutes, they had managed to rectify it. That is one.

The second, for example, is a system to monitor the collector shoes. Members will remember that this was the cause of the major disruptions back in 2011. There is a system now to monitor the health of these collector shoes on an ongoing basis.

These are the things that we continue to explore to try to make sure that they have real-time, including telemetric data that is being sent back, so that they can be analysed as well as, in certain instances, trigger rectification action on an urgent basis.

May I ask the Minister whether he will consider another metric of measuring delays? He did talk about the number of delays of more than five minutes. Would it not be better to look at the number of commuters affected multiplied by the number of minutes delayed because a delay that is 45 minutes in duration involving two major East-West station is much more significant in terms of commuter impact than one in an off-peak period for six minutes?

I did not give Members all the statistics because we measure greater than five minutes' delay, greater than 10 minutes' delay and greater than 30 minutes' delay. We also measure whether any incident affects more than 20,000 commuters. These are some of the different metrics.

The reason why I mentioned unplanned withdrawal rates is that if you have a train carrying 1,400 passengers during the peak period, and if you withdraw the train, it may take you a couple of minutes because that is almost as quickly as they can do it, but it will take quite a long time before those thousand over passengers continue on their journey. That is something that is harder to quantify in terms of impact. To me, it is an indicator that if we do not bring down the withdrawal rates, then we do run the possibility that it will one day happen during the peak hour and affect far more commuters than it would during, say, the middle of the afternoon.

Page: 32