SPADs

A lot of the difficulty nowadays is there’s few people to ask if one isn’t sure about something. Service control, who *do* have the knowledge, exist in their own silo and have little or no mechanism to pass it on outside of their domain. The various changes to the structure of the trains side over the last decade have really not helped at all, not helped by progressive retirement of knowledgable staff.

As an example, service control TBTC training takes seven weeks, trains TBTC training is a few days (forget the exact figure). See the difference?! Obviously train staff don’t need to know the ins and outs of what the signaller does, but it’s still quite a polar difference in training time. I/Os and managers don’t get anything extra.

I think the Bakerloo and Picc will struggle within the next few years, with LU’s “sooner ATO comes the better” mentality not applying here for the foreseeable future. The Bakerloo perhaps have the benefit of being a smaller line so may hold up better, but I definitely think the Picc is going o go through a very rough patch in the medium term in the absence of some management focus on the right things.

In this case, timing relates to the speed of your train, rather than anything else. We're not talking about time interval working here!

There are certain assumptions and reliances within LU signalling designs which may require timing functions. One is related to the calculation of overlaps - the train performance data for each inter-station run assumes a train is starting from a known speed (usually 35km/h) and to enforce that a timer is added to the controls of the starting signal. In addition to the track circuits being clear up to the next signal and its overlap, you also require the platform track circuit(s) occiped for a specific time which proves that the train speed has been suitably reduced.

The same principle applies nearing junctions, where signal spacing and positioning is optimised for getting a train as close to the junction as possible so that you can get it moving again quickly when the junction is clear.

In this case you'd provide a signal at a full speed overlap away from the junction, and another with a shorter overlap nearer to the junction. A timer would then be provided on the first signal, and the speed of the timing section would be the speed of the overlap on the next signal (or what could be achieved if the train reaccelerates).

Similar rules apply to speed controlled trainstops in terminal platforms and sidings.

A few things on terminology: Approach control is the inclusion of a track circuit on the approach to the signal which has to be occupied for the signal to clear (assuming all other conditions are met). Notting Hill Gate inner rail is a good example.

Speed control is where you must operate a timer, and generally comes as either a one shot or a two shot timing section. One shot is literally that - you have one go, and the timer generally runs to the trainstop. Platform starters and speed controlled trainstops are examples of one shot timers.

Two shot timers give you one chance which is the way it should operate normally, and a second in the event that you fail the first. The second shot is usually a lot more restrictive and will pretty much bring a train to a stand before it allows the signal to clear. Where you have a two-shot timing section it's quite common to delay the clearance of the signal until the train has reached the point at which the second shot begins, so that it can't be compromised by a train which satisfies the first shot, then holds back and takes a run up to the signal.

Does that help, or have I just made things infinitely more complicated?

Agreed, surely there has to be differentiation between a simple accident / misjudgement and reckless behaviour.

I sort of agree, however neither of these scenarios are actually common. Most SPADs take two forms - either a “start up against” where the driver simply fails to check the starting signal and the first awareness is when the train gets tripped, or a delayed reaction to a running signal (for one of a number of common reasons) followed by a swear-word moment at the last minute, normally leading to a last-minute emergency brake followed by an overrun of typically a car’s length.

It’s very rare for a driver to make no reaction to a signal at all, except in the case of starting signals.

Any SPAD is hazardous, as even if we work on the basis that the trainstop / overlap will almost certainly work as designed, following the SPAD the train is now into degraded working territory - instead of being protected by the signalling system and its designers, the liability for the safe separation of trains shifts to operating staff, normally the train operator and signal operator, but possibly including others such as station staff, operating officials or other control staff. Straight away a risk is introduced which wouldn’t have otherwise existed. There’s plenty of occasions where risk has bitten, simply because humans make mistakes.

Interesting responses to my original point. Thank you all.

Perhaps with SPADs, we are forced to look at the lowest common denominator and assume worst case but there should be some sort of understanding about driving skills. If you take the consistent braking and acceleration of the Victoria Line, no one complains about comfort apart from crowding. On a manually driven line, the range of driving performance is and always has been very wide. I recall being on a Picc train recently where the driver knew exactly how to get a full brake, consistently at every station and get the doors opening just as the wheels stopped. He (I saw him, he wasn't a she) saved several seconds at each station. If the fear of SPADs prevents this sort of skill, the driver's probably in the wrong job.

A semi-related question. While waiting for a Met train yesterday at Finchley Road, I noticed the northbound starter has co-acting signals that are not directly opposite each other (the one on the left side of the track is a few feet further away from the platform than the gantry-mounted one is). This made me wonder whether stopping between the two signals (when they're displaying danger) would be classed as a SPAD or whether both signals need to be passed for that?

Pardon?!

Things have changed over the years. When I started as a driver 14 years ago we'd visit the signal cabins as part of the initial "hiking" and the Instructor Operator would deliberately turn a blind eye whilst the signaller explained that it was best to contact them on a non recorded line in a SPAD scenario. Indeed the culture around SPADs at auto's was very blasse back then too.

Fast forward to today and we have Trackernet which everyone can see (yes, I know it can't be trusted), trains with black boxes and a radio system that tells the controller when the train is placed in a restricted mode. You simply cannot "get away with it" any more.

LU's discipline system is also much more rigorous these days.

It's a different world now and whether your opinion matches it or not, SPADs are dealt with as a safety critical error these days. As a result defensive driving techniques are a must on a manually driven line as a tool to help avoid commiting one in the first place. Chasing reds is the best way to get one's self put back onto stations!



Yep, the co-actor will have been provided because sight lines are different between A stock and S stock.

Co-actors are usually only provided where trains usually come to a complete stop - generally they're at station starters but there are some out on the running lines. As soon as the front of a given train passes the first signal, a SPAD has occured. There's no leeway bewteen the two, though in any case they are normally never more than a few feet apart.

(Perhaps this should be split off?)
Turn up early and wait; turn up late and you get chewed out. Pretty much guarantees more of the former. There's also the question of how long the waits are- 10 seconds is one thing, 60 is quite another.

I suggest that timing points have little to no meaning to the average rider. Few will know of them, and other than first/last trains, it's unlikely anyone is going to time their travels to them*. And when a train sits in the station for no apparent reason they/we start to wonder and "keeping to time" isn't high on the list of reasons.

*this isn't the same as "leave the shop at xx:yy to catch the next train"

Sorry, but I have to ask: What is 'chasing reds'?

Driving up to them as fast (too fast) as possible, perhaps?

Driving along a section of track fast in the hope of catching a red so you can tie your shoelace/redo your nail varnish?

Thats correct, it is a repeater behind the actual signal, if it was before it it would show a yellow.  Usually for sighting purposes and I can only think of one instance which is not at a station starter.

Driving rapidly such that you are continuously seeing red signals that then change to a proceed aspect ahead of you. This will get the most capacity out of the signalling system, but there is a higher risk with this driving style that as you hurtle towards a red signal expecting it to change (just like all the other ones you've seen are) it won't and you'll then SPAD. Driving defensively is a style of driving that I was taught on my motorcycle, it's always assuming the worst might happen so having a plan to mitigate it (eg in this context it means planning to stop a slight distance from a signal just in case the brakes aren't as good as you they could be).

Oddly, whilst the S7 enabling project put in a lot of co-acting signals, St James’s Park wasn’t one of them. At that site there was sufficient space to relocate the signal onto the platform, and a rather neat solution it was too.

It’s worth clarifying that Co-Acting signals aren’t repeaters; they display exactly the same aspects as the main signal, and in some cases the co-actor is the one you pass first. In any case, as Colin says, they both have exactly the same authority and passing either of them counts as a SPAD. Indeed, some of them have a track circuit between the Co-Acting signal and the main signal and that track circuit is included in the aspect control of the signal. It is usually proved unoccupied at time of clearance, and then cut out by another function such as the signal clear, so that a Train Operator doesn’t see it return to danger in their face. But if you draw up to the main signal and it doesn’t clear; it won’t, and you’ve had a SPAD.

In one case the offset between co-actor and blockjoint was such that Train Operators saw the co-actor return to danger through the cab door window - we had to fix this with a suitably positioned hood to shield the red aspect as any circuit-based solution wouldn't work.

But most of the passengers aren't railway people. We expect the trains to move, not to sit; any wait is seen as a delay regardless of what's announced. (As also seen in bank or grocery store lines.) (I don't think I've ever heard "We're running early, so need to wait a minute before leaving." but my sample size is quite small.)

This gets back to regulating the speed so there aren't long "delays" at the stations.

There are many cab-ride videos on Youtube of New York subway trains, and on many of them you can see 'chasing reds' happening.