2009 Q2 query- proving within MA or proceed aspect

[peternrz]

"The proving of a route prior to issuing a movement authority(or proceed aspect)"-how do I draw the line between discussing movement authority used in multi aspect signalling and that used in communication based signalling eg;in moving block?

[Peter]

"The proving of a route prior to issuing a movement authority(or proceed aspect)"-how do I draw the line between discussing movement authority used in multi aspect signalling and that used in communication based signalling eg;in moving block?

I don't think the question is looking for a comparison between the two things - both are mentioned to give you the option of answering the question based on the practice and systems with which you are familiar. In other words, pick one, say what you have picked and answer for that system and the way your railway uses it.

[dorothy.pipet]

An attempt for comments please.
While I wrote this out in about half an hour, I had been thinking it over for quite a while before getting round to it.

[PJW]

Not looked at detail yet and might be day or so before I do, but I think you did well to utilise the presentation that you chose.

I think you interpreted "infrastructure" a bit narrowly; I'd have included an example of all the basic components of an aspect level, so therefore proving that the railway had no trains on it and that there was a defined end of the movement authority (probably both a visual means of knowing where this was phsically located- say an illuminated signal- and a way of intervening to ensure that train stops within the associated safety distance beyond that- say an active TPWS loop.

Like other IRSE exam questions, if you are good you will not run out of things to say in 30mins, so need to make sure that you cover the whole width of the question rather than concentrate on only part of it because you have the depth of knowledge that permits you to do so. Detail is good in at least some areas, but prioritise breadth of coverage.

I do not think that you absolutely have to put something in every box, but what about the need to stop traiins running whilst a crane is installing a bridge section (or indeed over lane message sign gantry) for the motorway running parallel and adjacent to the railway. Perhaps unlikely to be implemnted in circuits as not going to be a regular scenario presumably, but perhaos the "inclusion" could be purely procedural, with the signaller utilising a reminder appliance over the entrance buttons etc. ........

An attempt for comments please.
While I wrote this out in about half an hour, I had been thinking it over for quite a while before getting round to it.

[PJW]

Additional to my previous comments:

Being pedantic, the question was about "the proving of route" and thus meaning controls in the aspect level continual evaluation rather then those elements only proved at the instant when it is decided whether appropriate for route to set.

Hence for points in the route, I'd have answered emphasising that what would be proved is that the points are in the state of having been set and locked (as opposed to doing the settting and doing the locking).
[Just in case in your day job you are thinking of a fringe to E10k Western Region relay room, it is worth you knowing that the WR's WZR is the "set" relay with two states: Normal and Reverse (BR936) as per a lever in a mechanical lever frame and there is a separate LR "Lock Relay". The LR is proved de-energised in all the aspects reading over the points (and its front contacts are double cut into the circuit for the point contactors).
Hence in these circuits the "proved set" is fundamentally the NWZR or RWZR (green tick on the Control Table when testing using the former standard B35) and the "proved locked" is the back contact of the LR in the aspect (blue tick on the Control Tables, ditto). Thus on the Test Copy of Control Tables you will expect to find a set of 3 ticks through each point of a route (the red tick is the detection itself).

May have been worth using the phrase "safe wheeled path" rather than "route" as it is what distinguishes from overlap and flank.

May have been worth you stating WHY the latest NR standard is not to detect points in overlap (basically "risk based"- the fact that a point detection failure only holds the immediately protecting signal to danger but not the outer one means less handsignalling, less errors and hence safety gain by omitting is judgefd to exceed the apparent safety from inclusion- what are the chances of a SPAD occuring at the time that there is a point failure that would be capable of derailing a train?).

For flank points, worth stating why the point detection of the flank end may fail after having become established- generally because there is a train on the parallel line passing over the flank end (obviously if points are correctly adjusted and well P'Way maintained shouldn't lose detection, but in the real world it does happen occasionally). If detection included "neat" then the other train's aspect sequence will "blip" and hence that driver may see, ought to assume the worst and therefore stop to report to signaller. Detecting at time of clearance only prevents such reversions.

Re moveable bridges then think that detection is essential; the risk profile is somewhat different between a SPAD train encountering a set of facing points which is almost certainly in one lie or the other even if the switch rail isn't quite hard enough against the stock rail (at slow speed pretty unlikely to derail- even if it does train may well stay broadly upright and will soon some to a stand) and the complete absence of bridge that may be swung diagonally across the river, which is likely to end up with a large splash and several vehicles of the train descending into the water. Likelihood low, but consequences extremely high and therefore the riisk is greater.
Certainly there are no swinging overlaps over swing bridges, or lifting overlps on lifing bridges!


Ground Frames-
You didn't really explain that point detection of GF points is inherent in proving the release Normal (ok for mechanical points there may not actually be electical detection it is only via the mechanical linkage to the drive and whee provided the FPL, but in other cases then detection is proved at least at time of normalisation if not continually).I think however that I would have lumped GF within points and would have used the time saved for something more different. I mentioned before that I interpreted question more widely than you, to include train detection, end of Movement Authority etc. However if you choose to stay with real "infrastructure" then think about structure gauge and tunnels etc.

#1 A car / lorry Shuttle train emerging from the Channel Tunnel must be routed into the Cheriton terminal rather than towards London because it is too large to fit safely. Similarly you would not want to be on a District line train routed by mistake onto the Picadilly line when approaching Earl's Court from the west as it would end up being like a cork in a bottle mouth.

#2 Also think about incompatibility of signalling systems- wouldn't want a non-ETCS fitted train to be signalled onto the Cambrian line because there are no lineside signals.

#3 In the future a London Overground train at Gunnersbury misrouted onto the District line would end up on an infrastructure without lineside signals or even train detection and thus it would disappear. Hence to protect the infrastructure beyond there sometimes need to be appropriate signalling controls.

#4 I don't think that it is often done, but certainly could consider traction power incompatibility. Certainly London Underground use Rail gap Indicators (a form of additional signal) that are normally not illuminated but light red if the traction power in next section is off. The assumption is that it must be off for a reason and so certainly wouldn't want a train to straddle the division from a live section and power up the dead section.
When that line is in future converted to CBTC operation the RGIs will be removed but the input used to remove the Movement Authority for a CBTC train sufficiently on approach to the section that train will come to stand before reaching the dead section; indeed it will also be added as an interlocking control into the colour light signals being provided for the Chiltern trains (even though these are DMUs and the risk of falsely powering up the section won't exist- there is still the risk to staff trackside that on an elctric railway consider themselves protected in an emergency situation when the power has been turned off).

It would be worth including something like the above; also explaining WHY in terms of assessing the risks, be they safety ,or to system reliability etc.
Note the subtle wording of the question:
Route = "expect", "reasons"
Beyond = "could be included", "reasons for inclusion or exclusion"
Adjacent = "may need to be proved", "reasons for consideration".

I don't think you really responded to that degree of finesse- yes you gave some reasons but did not reflect the decreasing certainty regarding the "correct" thing to do, trading the various advantages and disadvantages (often an apparent safety advantage is offset by the disadvantages of cost and particularly unreliability). Hence why the standards adopted by railway1 often differ from railway2 which has a different mix of traffic and environmental conditions; also why the standards of any railway may change over time as the situation and technology changes. NR provides TPWS to mitigate junction collisions but not generally plain line end on collisions; LU however provides train stops (or some better form of train protection) at all signals. I think the examiners were really looking for a bit more discussion of what drives such policy decisions, what are to perceived risks etc. before they would award very good marks.

Level crossing
Note that the barrier detection is non replacing in the signal aspect- it is only an up call on the barrieras which destroys the crossing clear. This is to allow for wind, people leaning on / attempting to lift the barriers etc; the likelihood is that the signals will only just have been cleared over the level crossing by the time the train gets to the one at which it would otherwise have to start braking to stop prior to the road, hence little value in replacing the aspect sequence subsequentlly if conditions change- the train is still going to be going over that level crossing since wouldn't have time to brake.

Regarding a "level crossing" adjacent to a route, perhaps you would allow this to be stretched to permit discusssion of a platform screen door. For a line such as the newer section of the Jubilee that has them, they are proved detected closed before a train can be signalled into or out of a platform.

If you really want true example of a level crossing adjacent to a route, the only one I can think of is just ouside Aberytwth. The Cambrian line passes over a ring road- I think it was an AOC(L) but now probably an ABCL but might be an AOC(L)+B. Anyway about 50-100m further along the same road there is another level crossing which is where the Vale of Reidol narrow gauge railway crosses that road before the lines converge and eventually enter the same station a mile or so away.

To avoid trapping motorists in the short stretch of road between the two railways and causing a tailback over the other level crossing, there is a procedural arrangement that the trains will not cross the road at the same time; the V of R train crew always contact the NR signaller to check where the Cambrian ATW train is before operating their level crossing. Not quite proof in the MA of a level crossing that is adjacent to the line, but the best I know and may be as good as it gets- unless of course you know different ......
However the task is not to rty to find some esoteric entry for each of the boxes in your grig; perfectly ok to have a line of entry as per yoour hazard detection trip wire

So overall not a bad attempt.
I do think it was too narrowly focussed, a but too much WHAT and not enough discussion as to WHY and the FACTORS that might influence the why.
It was clearly presented (I did think the tabular approach worked well even if it seemed you to try too hard to fill each one of the boxes which, in this particular question, was not necessary / appropriate in some instances) and, apart from a few quibbles, accurate. Definitely a comfortable pass; possibly a bit better than that because what was done, was done well.

However not convinced that it really got to the heart of what the examiners were after, so I doubt whether it would have been judged worthy of a Credit- if I am correct in this assessment then I think that it may have limited the marks they would be prepared to award; i.e. yu could convince the to give you all the marks earmarked for one element of the question, but there was little in you answer to cause them to give much out of the pot earmarked for discussing the consideration of whether or not suc additional controls should be provided.
I believe that you have "answered the question and nothing but the question", but not "the whole question".

An attempt for comments please.
While I wrote this out in about half an hour, I had been thinking it over for quite a while before getting round to it.

[asrisaku]

Hi again,

As a non UK candidate, it is difficult for me to write it long and also I have a weak background and most of study is self learning. I would like to share my attempt and it may be useful for someone to see bad and good things. This question, I had a spectacular friend review it.

Any feedback, please let me know.

Thanks
Arnut

[dorothy.pipet]

It would be best to start by saying which railway's practice you're using, suggest you write that at the start of every question's answer in the exams.

You've put a reasonable amount of material in your lists but very little for reasons/discussion. Note what PJW said of my attempt:
Note the subtle wording of the question:
Route = "expect", "reasons"
Beyond = "could be included", "reasons for inclusion or exclusion"
Adjacent = "may need to be proved", "reasons for consideration".

When you plan your answer, keep in mind that at least half the marks are going to be for the reasons. We sometimes fail to state the obvious, such as in the first part that [tracks clear and points set, locked detected] are in order to provide a safe wheeled path.

First section - instead of your note on the aspect sequence, it might have been better to comment on having a defined end of MA i.e. next signal lamp proved. I disagree slightly with your friend's comments on foul tracks - would be better in the third part.

Second section - I think a discussion on why different classes of route have overlaps, and why we use the lengths we do is expected, plus points in the overlap - whether they are locked or allowed to swing, and why they are/are not detected.

Third section - I think your friend is right that CAR is not relevant, but put in flank protection and foul tracks here. Overrun controls also fit, I came up with trip wires as another example.

[asrisaku]

Thanks Dorothy for these techniques. I will write it in the exam "UK Mainline practice" and think of the mark allocation, half will go for reasons.