Signalling a minories set in ex LNER BR Steam days.

[Nick C]

I'd say 2+28 would be on one lever, as would 9+22 and 24+6. No need for an FPL for ECS moves, only those carrying passengers. I'd expect the inbound signals to be to the right (higher numbers) of the frame with the points generally in the middle, and the distant and outer home to be at the rightmost end.

[Jeremy Cumberland]

49 minutes ago, Blackvault said:

Does 22 not need an FPL when propelling ECS back into P1 - P3? 

No. FPLs are only needed for passenger trains.

 

Several sets of points will be worked as pairs off one lever. I think your layout is pretty unambiguous with the pairs being

2 + 29

7 + 25

10 + 22

but perhaps someone else will propose an alternative.

 

I would place 5, 4, 3, 11 (or 3, 4, 5, 11) at the left hand end of the frame.

 

8 and 23 probably belong together, perhaps with 21 immediately to the right.

 

12 to 20 are probably numbered backwards. The distant will most likely be the furthest right lever (number 30, perhaps), the outer home will be next to the left, and the three main platform arms will come next. I don't know where the levers for the subsidiary arms would go (do you need subsidiaries? - you won't be shunting on the arrival line, so unless you are combining trains or have light engines arriving via the main arrivals line, I can't see any need for them).

 

@Nick C beat me to it - just.

Edited July 4 by Jeremy Cumberland

[St Enodoc]

23 hours ago, Jeremy Cumberland said:

I now have visions of someone building a tiny mechanism to place toy gun caps on the rail to mimic a detonator. I wonder if an 00 scale locomotive would be heavy enough to set one off.

Like this?

 

http://www.tri-angrailways.org.uk/fog.html

[Michael Hodgson]

On 04/07/2024 at 12:24, Jeremy Cumberland said:

I now have visions of someone building a tiny mechanism to place toy gun caps on the rail to mimic a detonator. I wonder if an 00 scale locomotive would be heavy enough to set one off.

I have seen that done.   Using one of those bomb toys they used to put in cereal packets, the loco striking a small arm  as it passes, the mechanism releasing the "bomb" to drop onto a wooden floor.  Only holds a single shot before you need to reload.  Crude but it works in O gauge; I don't doubt it could also be done in OO.

[The Stationmaster]

9 hours ago, Michael Hodgson said:

I have seen that done.   Using one of those bomb toys they used to put in cereal packets, the loco striking a small arm  as it passes, the mechanism releasing the "bomb" to drop onto a wooden floor.  Only holds a single shot before you need to reload.  Crude but it works in O gauge; I don't doubt it could also be done in OO.

Good job it's not on the track - you always used to get a noticeable impression of a  'thud' when going over an exploding det.

[Blackvault]

On 03/07/2024 at 11:11, Michael Hodgson said:

 

I'm not really keen on omitting the Home signal for arrivals into the platform.  If the back-scene break is a bridge rather than a tunnel, I think that would be correct (though the shunt disc would probably be on the far side of the bridge too), so you can still justify the omission if you want, by using slighly different scenics.  I don't think it would have a normal arm and call-on for the goods road as you've shown - I think it would just be a short arm.

 

The Up (Inner) Home signal would usually be located just short of the points it protects.  The concern would be that you don't want the risk of a shunt move being possible from the Down Main whilst an Up Train is in the tunnel (having already passed the signal).  This could be prevented using track circuits, but if it's a tunnel I'd still expect the splitting Home be there, probably repeated at the approach side of the tunnel.

 

I think there would also be  advanced starting signal just before the tunnel mouth if it's a tunnel, (but it would be off-scene as you've shown it if we've got a bridge here).

 

Thinking about this a bit more, I might change to an overbridge.  In that case, would I use a coacting signal(s) with the taller signals higher than the overbridge so that the inbound trains could see them? I assume all the calling-on signals would be at normal height as the inbound trains would stop before receiving the call-on signal?

 

I would also assume the advanced starting signal would be similar, not only for the drivers but the signal box being able to see all the signals it controls

Edited July 15 by Blackvault

[Michael Hodgson]

1 hour ago, Blackvault said:

 

I would also assume the advanced starting signal would be similar, not only for the drivers but the signal box being able to see all the signals it controls

Co-acting signals are for driver's ability to see the signal.  They would be quite a bit more expensive than siting a simpler signal on the other side of a bridge.  Co-acting tends to be used where curvature, gradients, overhanging platform canopies etc make it difficult to find a better site, and are more common when speeds are higher and the driver needs more room if he does have to stop.  Perhaps more popular on the LNWR than other companies.  Any subsidiary arms were generally not duplicated.

 

No requirement for signal to visible from the box - as long as there is an indication in the box - usually in the form of an electrical repeater, driven by contacts connected to the arm to prove the position.  These were quite common anyway, because you can't see much from the box window when it's foggy.

 

[Blackvault]

15 minutes ago, Michael Hodgson said:

Co-acting signals are for driver's ability to see the signal.  They would be quite a bit more expensive than siting a simpler signal on the other side of a bridge.  Co-acting tends to be used where curvature, gradients, overhanging platform canopies etc make it difficult to find a better site, and are more common when speeds are higher and the driver needs more room if he does have to stop.  Perhaps more popular on the LNWR than other companies.  Any subsidiary arms were generally not duplicated.

 

No requirement for signal to visible from the box - as long as there is an indication in the box - usually in the form of an electrical repeater, driven by contacts connected to the arm to prove the position.  These were quite common anyway, because you can't see much from the box window when it's foggy.

 

 

Would you therefore think, I could get away with my inbound multiheaded signal between the fiddle yard and the overbridge?

 

I hadn't thought of the view from the signal box on a foggy day.  Would my placement of the box be prototypical then? Speaking of boxes, how far up the line would the next one be? Assuming this is the suburbs of London? Would any of my signals need distant attached to them, in particular the outer home, outbound?

[Michael Hodgson]

The box placement looks ideal - close to the pointwork, so no excessive rodding runs, good view a view as possible of all movements, able to see tail lamps on all trains, no level crossing or single line token handling to worry about

 

I think in practice for the sake of a few yards the inbound signal would have been on the approach side of the bridge.

 

The advanced starter should be far enough past all point work to enable any likely shunting movement on that road to take place without passing it.  Although most of these may only be loco to/from shed, you may need to move a whole train from one platform to another, so that means at least a full train length.

 

Next signal box will be at next station as a general rule.  In depopulated areas that might be a long way, and you could need an intermediate or "break section" box in between in order get sufficient line capacity on a busy route.  However in a city it won't be far to the next suburban station, or to some group of sidings serving a factory, colliery etc or perhaps a level crossing.  But you don't want any more than necessary as each one adds about 4 employees (3 shifts + relief cover) to the payroll cost.

 

How far out the distant needs to be depends on several factors, mainly line speed (low in cities) and gradients - you need more space to stop going downhill.  On a hypothetical model with the next station is off-scene, it's entirely up to you.  You can deem the next box to be far enough to have its own independent distant, or closer if you do want to provide the extra arm.

Edited July 15 by Michael Hodgson

[St Enodoc]

7 hours ago, Michael Hodgson said:

each one adds about 4 employees (3 shifts + relief cover)

In the M&E it was a rule of thumb that for full 24/7 cover you needed five real people per position. You could just about get away with 4, plus overtime, but holidays and shift changes could catch you out.

[5BarVT]

7 hours ago, St Enodoc said:

In the M&E it was a rule of thumb that for full 24/7 cover you needed five real people per position. You could just about get away with 4, plus overtime, but holidays and shift changes could catch you out.

Back in the 5x8hr shifts per week days you needed to cover 21 shifts per week.  3 x shift technicians plus 1 x RDR (rest day relief) gave you 20 shifts.  A short turn round (8hr) effectively sneaked another one.  That gave nothing spare for holidays, bank holidays, sickness, training etc hence GPR to cover those.  In practice, it was more complicated as by my time it wasn’t a 40hr week.  But Sundays as Rostered Overtime gave some flex.

But as a rule of thumb 5 posts for 3 shift cover was pretty standard across S&T and (I think) operations too.

Paul.

[Blackvault]

16 hours ago, Michael Hodgson said:

The box placement looks ideal - close to the pointwork, so no excessive rodding runs, good view a view as possible of all movements, able to see tail lamps on all trains, no level crossing or single line token handling to worry about

 

I think in practice for the sake of a few yards the inbound signal would have been on the approach side of the bridge.

 

The advanced starter should be far enough past all point work to enable any likely shunting movement on that road to take place without passing it.  Although most of these may only be loco to/from shed, you may need to move a whole train from one platform to another, so that means at least a full train length.

 

Next signal box will be at next station as a general rule.  In depopulated areas that might be a long way, and you could need an intermediate or "break section" box in between in order get sufficient line capacity on a busy route.  However in a city it won't be far to the next suburban station, or to some group of sidings serving a factory, colliery etc or perhaps a level crossing.  But you don't want any more than necessary as each one adds about 4 employees (3 shifts + relief cover) to the payroll cost.

 

How far out the distant needs to be depends on several factors, mainly line speed (low in cities) and gradients - you need more space to stop going downhill.  On a hypothetical model with the next station is off-scene, it's entirely up to you.  You can deem the next box to be far enough to have its own independent distant, or closer if you do want to provide the extra arm.

 

Thanks for that helpful information. It is a shame that the inbound signal would be on the other side of the bridge (the grey rectangle) as it would add visual as shown currently, and hopefully operating interest.

 

As you can see I've altered the track work somewhat, however, it remains similar. I've placed Hornby signals to mimic the position and are not representative of the actual type of signal needed. As seen in the fiddle yard, the Advanced Starter would be mid-way down it, assuming that 5 coaches plus loco (shunter) are being drawn out of platform 1 and placed into 2.  The coaches used to represent this is MK1s, along with a loco, and while I intend to run the shorter '57 suburban stock and tank engines, I think using the longer coaches to represent this shunting move is the best way. 

 

I'll post in the correct forum for discussions about the track plan and the alterations from the original.

 

[The Stationmaster]

10 hours ago, St Enodoc said:

In the M&E it was a rule of thumb that for full 24/7 cover you needed five real people per position. You could just about get away with 4, plus overtime, but holidays and shift changes could catch you out.

Signal box staffing post introduction of the 40 hour week. was calculated as follows -

1 regular post per shift for a single manned 'box, i.e. 2 Signalmen for a two shift box, 3 for a 3 shift 'box

1 x Rest Day Relief for every 5 regular Signalmen posts working a 6 day week-  thus an RDR would cover more than one 'box.

1 x General Purpose Relief for every 8 posts (the 8 would include RDR as well as Regular men).   The purpose of the GPR was to cover leave and sickness plus short term vacancies and Groundsman/ some HandSignalling jobs but they were invariably also used to cover long term vacancies of both regular and RDR posts.

 

The proportions for Relief staff, particularly RD, had changed with the reduction to a 40 hour week during the 1950s, there was no further change until the introduction, where feasible, of 10 hour regular shifts in the 1970s - they had possibly  even appeared in some places in the late 1960s under a local arrangement.  Different numbers applied in the case of Crossing Keepers and depended on how a crossing was manned.

 

If all shifts at a 'box did not cover a 6 day week (Monday to Saturday) that meant that one shift would have an automatic Rest Day so the relief requirement would be reduced.  The most common shift to be taken out was the Saturday night turn but in some cases even the Saturday late (14.00-22.00) turn might not be required.  Sundays were always covered by overtime payment and although rostered were not part of the 40 hour 'Guaranteed Week'.

 

Signal boxes were classified depending on a combination of Equipment Value (e.g  the number of levers or block instruments or telephones) and the Traffic Value (basically the number of movements, including shunts, per shift) - these were termed 'marks'.   As traffic levels fell during, particularly, the the late 1950s and'60s BR agreed to a standstill on the grading of existing 'boxes unless the Signalmen at that 'box requested a reassessments of the marks (which was obviously a rare event because re-marking would inevitably reduce the grade of just about every mechanical 'box on the network).  Marks were still taken as considered necessary but they did not affect the grading of the 'box  but it was common to find what had been Class 1 'boxes scoring no higher than Class 3 due to the decline in traffic

 

The number of posts in a 'box depended entirely on workload (i.e. the mark and might mean a Booking Boy in some cases or a second, or more. Signalman in others.  There was a way of getting round the provision of a booking Boy by introducing what was known as 'skeleton booking; which meant the full detail of every train was not entered in the Train Register Book but just ceratin selected columns with the others left blank.   Skeleton booking was quite common at various large mechanical 'boxes on the WR  good examples being Radyr Jcn (over 100 levers in the frame) in South Wales and Portobello Jcn within a mile or so of Paddington which controlled all the running and ECS lines. 

 

Minories as drawn here would probably have ranked no higher than Class 1 and would inevitably have been a single handed job although as it was on the ER there might be a Telegraph Lad on part of the day.  The workload would have dropped massively when steam was replaced by DMUs and the Telegraph Lad's post would have gone but the standstill would have held the grading of the 'box.

[Blackvault]

I was thinking about levers. The cross-over between the down and up lines, would be a signal lever, would the points in and out of P1 be as well?  That is the Righthand and Lefthand points.

[Jeremy Cumberland]

48 minutes ago, Blackvault said:

I was thinking about levers. The cross-over between the down and up lines, would be a signal lever, would the points in and out of P1 be as well?  That is the Righthand and Lefthand points.

Probably.

 

I suspect that all levers normal would be trains to depart platform 1 and arrive in platform 2 (but it might be for trains to arrive in platform 3 instead). Ignoring FPLs and signals, you'd then pull one lever for a train to arrive in platform 1, and one lever for a train to depart from platform 2.

[bécasse]

I remember being taught at The Grove that seven people were required to ensure that a single continuously manned post could be covered without resorting to overtime, the apparent "extras" being required to cover things like leave (including public holidays) and "standard" sickness, it being reckoned that a single person wouldn't, on average, actually work more than 220 rostered turns of duty a year.

 

It was interesting at Eurostar when new universal contracts were introduced which required flexible working for not less than 1.800 hours, and not more than 2.000 hours, per November-October year. I can still remember the puzzled look on (Personnel Director) Peter Whittaker's face when the senior members of the Commercial Team asked him what they should do when they ran up against the 2.000 hours ceiling, a situation that apparently hadn't been considered.

[St Enodoc]

3 hours ago, bécasse said:

I remember being taught at The Grove that seven people were required to ensure that a single continuously manned post could be covered without resorting to overtime

I wouldn't argue with that, without overtime. Five, with a bit of overtime to keep them interested, seemed to work most of the time though. If it was a team of supervisors, a bit of higher-grade duty from the shop floor from time to time was also a good way to evaluate who might be suitable for promotion in due course.

[The Stationmaster]

15 hours ago, bécasse said:

I remember being taught at The Grove that seven people were required to ensure that a single continuously manned post could be covered without resorting to overtime, the apparent "extras" being required to cover things like leave (including public holidays) and "standard" sickness, it being reckoned that a single person wouldn't, on average, actually work more than 220 rostered turns of duty a year.

 

It was interesting at Eurostar when new universal contracts were introduced which required flexible working for not less than 1.800 hours, and not more than 2.000 hours, per November-October year. I can still remember the puzzled look on (Personnel Director) Peter Whittaker's face when the senior members of the Commercial Team asked him what they should do when they ran up against the 2.000 hours ceiling, a situation that apparently hadn't been considered.

The 2,000 hours plus situation was covered when it arose - give 'em the money.

A number of the Group 1 Drivers went over the top although by then it was set figure, not a range.  in fact you couldn't calculate an establishment for many jobs, particularly traincrew, on a range so there had to be a set Annual Hours figure.  But by far and away the biggest problem for traincrew information was actually recording the hours as the system intended to do it (and umpteen other things) was a computing fail of the usual worst sort - poorly spec'd, development given to an inexperienced company, then umpteen cost variation invoices submitted by the software company.  

 

A few months after my arrival at Eurostar I was duly tasked to review where that project had got to and where it was capable of going.  It was quickly evident that without spending a lot more money there wasn't a chance of getting anywhere near the original spec and, even worse, if that money was spent there was no guarantee that the software developers could deliver what was required.  So the work was stopped and the contract cancelled on my recommendation and we took delivery of what had so far been completed - which actually could tot up hours worked and 'sort of' be used for rostering   but couldn't do anything else which had been in the original spec.

 

But in fact getting to a proper spec was exceedingly difficult and I looked at various systems already in use in the UK or being offered to potential UK customers.  I finally found a company which understood how the required functionality would need to be handled using relational databases.  But their proposal was rejected on cost grounds.  After  left a group of directors decided - after a trip to Saudia Arabia to see it working - to buy a system I had previously rejected as totally inadequate.  Of course they didn't ask the people who had to use it or be rostered using it and it was dumped after abput 6 months.  Strangely the Directors could have gone to Swindon to see the original version of that software where it was being used for rostering BA flight deck and cabin crew.  Presumably a trip to Saudi was more attractive than one to Swindon?   The company's proposal rejected on cost grounds would have achieved the full functionality that was needed and the developers actually understood what the spec meant. 

 

Incidentally the figure of 7 to cover 3 post sounds excessive.  On the Western we used a 2.8 divisor for Drivers' establishment calculation - u i.we. the number of Diagram Days was divided by 2.8 to give the required establishment to cover assuming 8 hour turns and a 40 hour week.  With rostered leave that meant ona full establishment that we never needed overtime or Rest Day Working to cover booked work and the nromal level of specials.  If a depot had an unusually high level of specials then 2.7 was used.

[bécasse]

It was all but fifty years ago that we were taught about the need for 7 people to cover a 24h/24 7j/7 post. We were being trained as implementors of the proposed new "territory" structure, the financial rational for which required the application of properly costed standard procedures across the whole network, so understanding how costs arose was a fundamental part of our work. We were initially surprised by the 7 figure although it was obvious that, with the salaried staff working week comprising 38 hours, one needed more than 4 people to cover a single post that had to manned 168 hours a week, but our tutors took us through it and, adding in leave, sickness (a standard norm of just 10 days), training and a few other items, quickly demonstrated that an establishment of more than 6 was required. What is true is that the requirement for multiple posts was more than 6 but less than 7 per post.

 

In the end, not long after the finish of our training in fact, the implementation of several trial single-location standardisation schemes demonstrated that the intended savings would not be achieved and the territory restructuring was cancelled. This rather surprised some of my colleagues but, having trained a physicist, I was aware that nature determines that chaos produces the most efficient flow and I wasn't that surprised to find that it applied to administrative systems as well. Nevertheless our training was very thorough and I often found it useful in later life, particularly the injunction to always stand back and make sure that you understand the broad picture.

[St Enodoc]

3 hours ago, bécasse said:

It was all but fifty years ago that we were taught about the need for 7 people to cover a 24h/24 7j/7 post. We were being trained as implementors of the proposed new "territory" structure, the financial rational for which required the application of properly costed standard procedures across the whole network, so understanding how costs arose was a fundamental part of our work. We were initially surprised by the 7 figure although it was obvious that, with the salaried staff working week comprising 38 hours, one needed more than 4 people to cover a single post that had to manned 168 hours a week, but our tutors took us through it and, adding in leave, sickness (a standard norm of just 10 days), training and a few other items, quickly demonstrated that an establishment of more than 6 was required. What is true is that the requirement for multiple posts was more than 6 but less than 7 per post.

 

In the end, not long after the finish of our training in fact, the implementation of several trial single-location standardisation schemes demonstrated that the intended savings would not be achieved and the territory restructuring was cancelled. This rather surprised some of my colleagues but, having trained a physicist, I was aware that nature determines that chaos produces the most efficient flow and I wasn't that surprised to find that it applied to administrative systems as well. Nevertheless our training was very thorough and I often found it useful in later life, particularly the injunction to always stand back and make sure that you understand the broad picture.

Not wishing to challenge the collective might of The Grove's industrial engineering team, but how were those seven bodies divided up, how many were on duty at a time and what did they actually do when more than one was on duty?

[The Stationmaster]

On 24/07/2024 at 20:29, bécasse said:

It was all but fifty years ago that we were taught about the need for 7 people to cover a 24h/24 7j/7 post. We were being trained as implementors of the proposed new "territory" structure, the financial rational for which required the application of properly costed standard procedures across the whole network, so understanding how costs arose was a fundamental part of our work. We were initially surprised by the 7 figure although it was obvious that, with the salaried staff working week comprising 38 hours, one needed more than 4 people to cover a single post that had to manned 168 hours a week, but our tutors took us through it and, adding in leave, sickness (a standard norm of just 10 days), training and a few other items, quickly demonstrated that an establishment of more than 6 was required. What is true is that the requirement for multiple posts was more than 6 but less than 7 per post.

 

In the end, not long after the finish of our training in fact, the implementation of several trial single-location standardisation schemes demonstrated that the intended savings would not be achieved and the territory restructuring was cancelled. This rather surprised some of my colleagues but, having trained a physicist, I was aware that nature determines that chaos produces the most efficient flow and I wasn't that surprised to find that it applied to administrative systems as well. Nevertheless our training was very thorough and I often found it useful in later life, particularly the injunction to always stand back and make sure that you understand the broad picture.

I wonder what rostering system that was based on?  With a 38 hour week covering three 8 hour shifts there is a need for Rest Day Relief to cover 6 more hours per week than is required with a 40 hour week.  This is readily achieved by putting an extra Rest Day into the roster cycle - in other words the regular staff still work 8 hour shifts and do 5 shifts per week but each person accumulates 2 hours per week to be covered by an additional Rest Day giving a total of 5 Rest Days in a 4 week roster cycle.  

 

The Rest Day Relief also gets a 5th Rest Day in the 4 week cycle and this what drives the cost increase resulting from reduced working hours as it means that overall the number of posts an RDR can cover will be slightly reduced.  General Purpose Relief would still work on the original basis of one per every 8 posts but that would vary depending on amount of leave and estimated sickness required to be covered.

 

But it still means that you only need a base of 3 people plus an RDR (actually part of an RDR) to cover the Guaranteed Week of 3 daily 8 hour shifts Monday - Saturday.  Which takes me back to my opening question.  If you work a traditional roster Sunday is outside the Guaranteed Week and the turns are covered by a combination of either 12 hour shifts or 8 hours shifts with the latter meaning a higher frequency of Sunday turns and more double backs on short (8 hour) rest.  

 

But if you include Sundays in the full rostered week and regard them as part of the Guaranteed Week you immediately create an additional 24 hours needing to be covered on the basic roster and you increase the shortfall of available hours to cover them - so you need more staff.   This is one of the sometimes missed implications of what used to be called on BR 'Continental Rostering' - the instant you make Sunday part of the Guaranteed week you need more people to cover the same amount of work.

 

Having over the years managed numerous supervisory and Control Office staff covering posts on a 3 shift (in most cases) 7 day per week basis the 5th Rest Day in a four week cycle was inevitably how we achieved 24 hour cover with people working a 38 hour week.   Deciding who got a share of the Sundays was sometimes a far bigger headache and source of argument - at one place I even had to put into the Local Agreements a definition of which day of the week Sunday was (i.e. was it the first or last day of the week?)