2013 Q1 - London Underground Style CTs

[AdrianM]

I have created a new thread to this question as I have attempted it using London Underground style principles.

Find attached my control table for part a).

If anyone is familiar with LU principles I would very much appreciate any input as I've obviously had to caveat my answer with several notes and assumptions. + I'd like to know if I've gone wrong somewhere!

I have also made effort to highlight all my acronyms and give a little explanation against certain terms.


While I'm reasonably happy with my control table, I'm less happy about what I'm to do about the route and lever locking....to do this correctly would take an awfully long time and without doing the entire site would be somewhat 'incorrect'. (a side note: How many marks would you anticipate I would receive for only the attached control table?)

Also, as LU don't generally don't use/have point control tables I'm a little lost on the actual content of such tables...other than simply stating any track locking tracks and the associated routes that use the point....is that really all that is in these tables?


If anyone has experience doing the exam LU style I welcome your input! I know there is only a week to go...I should have posted this months ago.....sigh....


Regards,
Adrian

[PJW]

I have created a new thread to this question as I have attempted it using London Underground style principles.

Find attached my control table for part a).

If anyone is familiar with LU principles I would very much appreciate any input as I've obviously had to caveat my answer with several notes and assumptions. + I'd like to know if I've gone wrong somewhere!

I have also made effort to highlight all my acronyms and give a little explanation against certain terms.


While I'm reasonably happy with my control table, I'm less happy about what I'm to do about the route and lever locking....to do this correctly would take an awfully long time and without doing the entire site would be somewhat 'incorrect'. (a side note: How many marks would you anticipate I would receive for only the attached control table?)

Also, as LU don't generally don't use/have point control tables I'm a little lost on the actual content of such tables...other than simply stating any track locking tracks and the associated routes that use the point....is that really all that is in these tables?


If anyone has experience doing the exam LU style I welcome your input! I know there is only a week to go...I should have posted this months ago.....sigh....


Regards,
Adrian

Yes you definitely should have much earlier!

Not really qualified to comment (my level of competence is no more than "vaguely aware", certainly not "familiar"), but in the absence of anyone else so far......

In my near ignorance, the entries overall seem reasonable for the ELECTRICAL VITAL controls.

However it does not seem to depict any
1) Route setting controls
2) Locking


I am well aware that 1 may be done by a non-vital system, possibly a computer, at the Control Centre
and 2 may be undertaken mechanically on the frame / interlocking machine.

• You definitely need to give details of the locking; I would interpret this as showing all the Locks and Releases on the lever for the route in question (reflecting the dogs and cats).
  
• I think you should at least cover in a generalised note (or perhaps discuss as an example for one route only)  the things which would be checked prior to applying compressed air to attempt to throw the levers; it is even more important for the traditional LU signalling than other varieties of signalling that the vital locking is not "stressed" by the non vital control system.  If the control room operator selects a route, then each of the point levers must first be moved and only when proved to be correct should the signal/route lever be operated or else the interlocking frame may become jammed.  Hence whereas LU may not regard this logic as a "Control Table" and perhaps is quite simple and inherently defined to be the same as the point controls within the signal's aspect, I do feel that somewhere you should explicitly state this and explain.
  

The wording of the question explicitly says that if your interlocking is part electrical/electronic and part mechanical then BOTH should be shown; hence I think if you failed to comply this would be viewed severely. You might get away with not doing much, but totally omit at your peril.........

[For ease of reference, I have included as an attachment showing the defined routes and points marked up on layout]

I think if you did a lever locking table which showed-

1. all the locking for each of the routes you have been asked to do (105A_M, 119C-C) 
   
2. with only the relevant reciprocal locking for each of the points locked by those routes (213, 214, 215, 218, 219, 220, 221)
   
3. all the locking for each of the points you have been asked to do (212)
   
4. with only the relevant reciprocal locking for each of the routes which lock those points (113A_S, 115A_S, 122A_M, 122B_M, 132/4/6/8 combined)
   

(and explaining this is the case for the partial entries for items 2 & 4) then this might be a reasonable interpretation.  
Given the fact that point control tables don't really exist (fundamentally because the locking is simple and basically done via the route) then the total effort expended would seem to be broadly equivalent to someone attempting to do Route & Aspect and Point CTs to NR standards.
Actually it isn't only LU for which Point CTs are not actually totally appropriate; the same can be said for Indian Railway practice where it is heavily based on mechanical signalling and lever frame interlocking.

The "Point Control Tables" would very largely be covered by the above presentation.  Given that it is rare that sectional route release is employed, then there is basically little else to show as those routes would be holding locked all their points until they themselves are normalised (as defined by the Approach Lock & Backlock entries). As you say it would only be a matter of listing the relevant deadlocking TCs over the point ends and any foul (or conditionally foul) TCs and I think you could cover by notes such as #1 on the point entry on that same locking table and then define #1 = Requires DJ clear N>R and R>N


The best advice I can give; it would be good though if someone from an LU background, particularly someone who has actually done such in the exam could comment

[ODP359]

PJW sent me an email with a link to this thread this evening - a nice surprise before a night shift!

If the site were laid out to LU practice I would not expect to see anything other than a reduced overlap on 119 signal with 105 signal requiring either 119 'off' or a timing section on the approach to 105 to validate the reduced O/L.

However, in terms of answering the questions as asked, I would expect the controls of 105 to include up to and including CH and mechnical locking to be 105 x 214.215
However I would then expect the overlap locking to be 'handed over' to 119, with the backlock controls being 105G and V 'on', and (CF occ, 119 ®) or (CF occ. CG clr for 15 seconds).

The second question relating to 119C © is somewhat harder to replicate in the LU world - as LU hasn't coupled on the main lines since the 1970s.
I would deliberately omit DA from the controls (this being the 4-car portion the second train is due to couple to) and likewise omit all the proving of the TETS trainstops and their associated timers (assuming that these would be operated by the first train already berthed in the platform). I would additionally expect a train at 119G to be proved not only at rest but also to be a short train suitable for coupling.

In terms of the mechanical locking, I do feel that 119 is somewhat complicated to mechanically lock quickly (It's 3am) and in reality would require some conditioning. Rather than looking directly at the lever locking I would first just list conflicting routes which must be normal. The three traditional non-vital circuit levels used by LU do this - in particular the UR which is used to feed lever motors. Before the UR is allowed to pick all conflicting URs must be down - which in itself is a form of non-vital locking between different routes at route level.

One observation I do have is that I think there is too much reliance based on the technology used by listing relay names and their positions, amking the control tables look more like circuits expressed in tabular format rather than tables of controls.
For example '200 VCR up'; why not 200V 'on' and in the Approach lock, how about '105 signal and trainstop on and CB clr'?

[PJW]

I have now acquired some more information from a third party which suggests:

1. The LU Signalling Design Handbook has a section on CTs which would be worth utilising as an Exam standard.

2. Control Tables are not always utilised but are often included in the bookwiring to supplement the written circuits , but these are sometimes regarded as sufficient by themselves to define the controls.  Apparently the Bakerloo and the East end of the District line are good places to look.
Tend to be regarded as more essential where interlocking is implemented by data or within relay interlockings; West Ham is suggested as a recent example worth referring to.

3. If the site is mechanical, LU always provides a locking table (in addition to the locking diagram which shows HOW the locking is implemented).  This covers route interlocking (i.e. signals and points), but not route locking (approach / backlock), track locking or proving. 

4. It is suggested that it is very possible to produce a locking table in exam conditions.  Recommendation to avoid point-point locking (unless wide to gauge traps) as this tends to lock up the site in failure conditions- may be regarded as the difference between a Pass and a Distinction answer.
State assumptions where it is possible to implement route locking either mechanically or electrically; for example back-to-back was traditionally done mechanically but nowadays the preference is just to do at aspect level to avoid excess disruption if a failure occurs.

5. It is suggested that LU actually do often see Control Tables as tabulated circuits, so details of the technology of implementation can be relevant in that context; in fault scenario there is a difference between 200VCR up and 200VPR down and this might be a rationale for showing as such rather than the more conceptual 200V "on".

As ever, put 3 signal engineers in a room and get 4 opinions!

[ODP359]

Now having had some sleep and a chance to review...

Firstly some errors in my original control tables; namely the omission of any trainstop proving, which would be done via the pick-up of CD and CG track circuits respectively.

119's should also require the timer associated with DB occ to be proved at zero as a miscellaneous control, and the requirement for CF occ for 15 seconds should really be CE clr. CF occ for 15 seconds to prove a short train suitable for coupling. Whoops!

Taking PJW's suggestions above, some commentary on the provision or omission of control tables. These were originally a separate drawing, not included in the wiring diagrams, and used to be provided simply for the RI submission. The 1950s resignallings tended to then develop them into what became 'selection schedules' - simply a circuit expressed in written format rather than drawn, and quite possibly done after the circuits themselves were finalised! Some of these have now not been kept up to date, others have been redrawn more as 'proper' control tables. Ealing Common for example has the selection schedules, and Watford South Junction has them redrawn more as a control table.

The 1970s non-safety circuits had a form of control table showing what routes had to be normal at a  non-vital level - mainly because the drawings themselves were massive and hard to read quickly in failure conditions.
The Bakerloo line had a rather comprehensive set of 'used in' and 'proved in' tables for treadles, timers etc, but never really any control tables - again they were done for the RI but never made it into the site book. Anything post 1991 tended to have control tables along the NR format, so the Central, Waterloo and City lines, West Ham, Heathrow T5 etc.

The pros and cons of point to point locking are still a subject of debate in the LU community; in my early days as a maintenance tech I worked on the Bakerloo which had very limited point to point locking for the reasons PJW states; the downside of this is that the signal to signal locking often has to be more complex with greater conditioning and occasionally has to use a 'route lever' (effectively a phantom signal) to simplify the mechanical locking arrangements. As ever, it often comes down to which design team produces the design.

On that note, I am shortly due to commission a mechanical locking alteration which when I first looked at it lacked point to point locking (it was never provided and the locking was achieved signal to signal), but the designers subsequently changed their minds and provided the point to point locking. I think a lot of it, rightly or wrongly, comes down to the preference of the approver at times.

If anyone wishes to look at a set of LU control tables as used for RI submissions of the past, a good example is on pg53 of Green Booklet No.19 - this also has an example of a point control table, which is somewhat basic! The Green Booklet is also quite good at laying out the general principles of the call (the USR), the route interlocking relay (UR), and the lever operation circuit, which also has to prove the lever positions and mimic the mechanical interlocking to prevent jamming. I'll try to get a scan in the next day or so.