System Design of North-South and East-West MRT Lines

Mdm Speaker, as explained at the Land Transport (LTA)-SMRT joint briefing on 29 July 2015, the disruption on 7 July on the North-South and East-West Lines (NSEWL) was caused by intermittent tripping of the power system at multiple locations. Chloride deposits, possibly exacerbated by a water leak in the tunnel, lowered the electrical resistance of a third rail insulator between Tanjong Pagar and Raffles Place stations. This allowed electricity to flow through the insulator to the ground and led to a higher-than-normal voltage difference between the running rail and the ground. Because the running rail is connected throughout NSEWL, the higher voltages would show up at multiple locations across the network. The movement of trains further elevated the voltages.

When confronted with abnormal voltage, the 64P safety mechanism would be activated as a safety precaution. This safety mechanism is widely used in overseas systems, such as Japan's as well as in all our rail lines. In this instance, the safety mechanism was activated to trip the power at multiple locations across the two lines.

Page: 23

As to whether it was design or maintenance flaws that precipitated the incident, I would say that LTA's investigation is ongoing and so it is perhaps premature to prejudge the conclusion. But, as a general rule, I would say that if there are design flaws, we are likely to see it pop up within three to five years or so of operating the line, rather than to see it come up 25 years or so after the start of operations.

In the interim, SMRT is replacing all the third rail insulators. SMRT has also combed through the tunnels to ensure that there are no other leaks with water dripping onto trackside installations and will be installing data loggers at all traction power substations within the next two months to better monitor the condition of the insulators. Lastly, we will be making NSEWL less susceptible to such power trips by increasing the touch voltage threshold for the 64P safety mechanism to levels similar to international standards, as well as the newer Circle and Downtown Lines.

Thank you, Madam. I have four supplementary questions for the Minister. First, given that NSEWL started operations in 1987, when did its rail assets, specifically the sleepers, signalling system and third rail reach their end of service life? Second, when did the renewal and replacement works for the sleepers, signalling system and third rail commence? Third, is it LTA or SMRT that is responsible for the renewal of these assets? In other words, who owns these assets? And who is supposed to initiate the process of renewal and who is supposed to bear the cost? And lastly, what was the cost of the renewal and replacement of these assets and how much of this cost has been borne by the Government so far?

If you look at the sleepers, for instance, and you are talking about the wooden sleepers, there were two sets of wooden sleepers that were used. One set had deteriorated more quickly than the other, even though it was installed later. Therefore, it is a little bit hard to give the Member a definitive timeframe as to how long these sleepers can last because it depends on their conditions, the type of wood, how it is treated and a variety of different factors.

Therefore, the more important thing is actually to monitor them on a consistent level and replace them on an ad hoc basis as and when necessary. But when it reaches what is considered the critical threshold, then you want to consider changing them out, replacing them en masse and, in this case, we are doing so by replacing them with concrete sleepers.

Page: 24

For the third rail, the programme is ongoing and we will be able to replace the third rail in its entirety by 2017. There is a slight delay to the programme because there was a need to try and de-conflict all the works that are ongoing, like replacing the sleepers which we give priority to, the re-signalling which we also give added priority to, over the third rail. But this is ongoing now and its replacement will complete in 2017.

On who is responsible for initiating the review, it is the operator's responsibility because they are responsible for the maintenance and, so, they will have the best idea of the condition of these different components.

As to who bears the cost, there is a formulation, quite a complicated one, in the original White Paper, which I do not intend to go into, but we had talked about this previously. Basically, the Government bears part of the cost of the replacement and renewal. But because we think that actually it could have lasted us longer if it was better maintained over the years, therefore, you also take a discount from what the Government would otherwise have to pay. But the formulas are all set out in the White Paper.

I believe I have answered the Member's questions.

I thank the Minister for the reply. I did some research on the SMRT website and, according to the frequently asked questions (FAQ) section, the timber sleepers are supposed to last between 15 and 25 years. Can the Minister verify that this is the case? Does the operator replace the sleepers only when it reaches its end-of-life or does it replace when it sees signs of wear and tear? Because if we count 15 to 25 years from 1987, the sleeper replacement ought to have been completed by 2002 or 2012. But according to SMRT, the sleeper replacement only started in November 2012 – this is the press statement that they issued – and it will not be completed until 2016. So, why was there such a long delay and who is responsible for this delay? And did the delay in sleeper replacements pose any risk to commuters' safety at any point?

Regarding the signalling system, again according to the SMRT website, the replacement of the signalling system started only recently and this is more than 28 years after the line operation started. My last question is: have there been any problems or conflicts between LTA and SMRT that resulted in delays in the renewals of these critical rail assets? And how has the Ministry of Transport (MOT) stepped in to resolve these differences so that there is no delay in the renewal process?

As I have said before on sleeper replacements, it is an ongoing process. The sleepers do not deteriorate to the same extent, even though they may be within the same locality and that is borne out by some of the pictures that I have seen over time. So,

Page: 25

you may have one or two sleepers within the section that have deteriorated more significantly than any of the others.

What we do is, on an ongoing basis, we will replace them. But at a certain point in time, when the assessment is that in order to be at the top of the curve, you would want to replace them en masse and with a different set altogether, in this case, with concrete sleepers. Were there any risks to commuters? The answer is no.

On the signalling system, actually, the signalling system is still working reasonably well. Part of the reason why we initiated the upgrading of the signalling system was because we wanted to be able to run the trains at shorter headways. The current signalling system allows trains to operate at 120 seconds. The newer systems, including that in the North East Line, for example, will allow you to run trains at tighter intervals. And we wanted to be able to do that in order to add capacity to NSEWL, and that was one of the primary reasons why the signalling upgrading project was initiated.

On conflicts and delays between the operator and the regulator, as I have explained before, it is actually a fine balance. If they have too cozy a relationship, you get worried; if they are at loggerheads all the time, you will get worried, too. So, we expect a certain healthy tension because one is operating it and one is regulating it, doing the audits, doing the checks, providing some of this feedback to urge the operator to take certain actions and to expedite what they are doing.

What they do is that they report to me on a three-monthly basis. After the December 2011 incident, I felt that the way LTA was overseeing the operators needed some improvement. In the past, we were looking at the outcomes and the indicators that we get every so often. But by the time the indicators take a clear turn for the worse, it probably is a little bit too late. That was number one.

Two is that because we were initiating a whole series of major projects, there was a need to make sure that there was proper coordination, both between the regulator and the operators as well as within the operator itself. I wanted to have more direct oversight of all these different arrangements. So, I initiated the joint teams and I required them to report to me every three months.

What we have done more recently is to step up the audits by LTA on the operators, both planned and surprise audits, and also to embed a dedicated team from LTA in SMRT so that we can have a better insight of what is happening, on an almost day-to-day basis.

Page: 26

Thank you, Madam. I appreciate the Minister's point that because there are so many replacement works that need to be done, there is a need to prioritise them. Would the Minister agree that a lot of these replacement works should actually have been started earlier on, pre-2012, in order to be able to complete the replacement works in time before the end-of-life of the system?

I think it is not always clear when the end-of-life is. There may be an estimate given but, because of varying conditions, that can always change. In any case, I suspect that perfect foresight is as rare as 20/20 hindsight is common.