Single Line Operation = Capacity Enhancement

[najaB]

Thanks for that........Bi-directional operation of loops is the norm on USA and other countries lines signalled with CTC...

I think it is generally the case in the UK as well for loops on single track lines.

Edit: RETB lines excluded, of course.

 

[MarkyT]

This means basically that two trains can't enter a passing station simultaneously unless the loop extends for at least 200m either side of the station. There are some situations where this can be reduced or avoided but they involve running the trains more slowly.

On branch lines, a high proportion of loops have shorter overlaps, and hence a shorter overall loop length, in conjunction with lower run-in speed, which is often necessitated by the loop turnouts anyway. An additional turnout leading to a short spur terminated (traditionally) by a sand drag or more likely a friction buffer stop today can also be used sometimes in lieu of a full overlap. A fast loop like Axminster clearly needs, and has full overlaps.

I think the Penrhyn arrangement only works if one train arrives before the other one.

Correct. See my animation of a Penryn style solution at Brading on the Isle of Wight.

https://youtu.be/1H9EV0iT0dQ

The design is most suitable on shorter self-contained branch lines. In my example if a first train was a little late arriving from Shanklin, the second train from Ryde would have to wait outside Brading until the first arrived. That in turn might add a little further delay to the first train departing but as long as there's sufficient terminal layover in the timetable at each end of the branch, services should be able to recover.

Another constraint for Penryn style is platform length and whether space is available for the extra long construction. Even with only very short trains the combined length of the sequential platforms adds up to 238m at Penryn itself. On many branch lines, that kind of length may not be available between other obstacles at either end like bridges or level crossings.

The passing loop outside the station also occurs to the east of Tisbury. The loop (both tracks) are bi-directional. If a down train is late the up train will leave the station and enter the loop. The down train will then pass on the through track and stop at the platform.

The opposite happens if the up train is late. The down will wait in the loop and proceed to the platform once the up train has cleared.

On my various journeys to Yeovil the latter seems to happen more often.

They didn't put this loop outside the station entirely by choice. It was a new facility added by BR in 1986 to split a previous unwieldy 19-mile single track section dating from the original 1960s simplification scheme. A two platform solution would have been far preferable, but because BR had previously and very shortsightedly sold land on the down side, including the platform right up to the former face, there was no space available without some no doubt very expensive and difficult reacquisition process, which BR was unable or unwilling to afford at the time.

Typically, it is up trains that are looped, and they are timed to wait for 5 minutes which is clearly added to their overall journey time as they also stop at the station just beforehand. From arrival at Tisbury to departure from the loop it takes 8 minutes. In the up direction Tisbury to Salisbury with the additional loop stop is timed for 19.5 minutes, whilst the down run from Salisbury to Tisbury and without a loop stop is timed at only 13 minutes. By contrast the dynamic looping operation at Axminster adds no time whatsoever to the journey in either direction.

A Penryn loop would be better option theoretically at Tisbury if buying back the former down platform remained impossible, but the short (former up) platform there is only capable of standing 3 cars today (SDO is used for anything longer). Wth the potential for up to 9 cars on this route a new combined Penryn style platform could end up a monster 500m or so long, almost certainly impossible to fit in the space available without further land purchase on the up side (see attached image).

 

[furnessvale]

I seem to recall that the section Altrincham to Navigation Road carries more trains now that it is operated as two bidirectional single tracks than it did when it was operated a one double track section. At a cost of greatly increased journey times between Northwich and Manchester.

It may carry more trains(trams) now than it used to do, but that is a far cry from claims that capacity has actually been increased by the conversion.

In the past, it was simply a grossly underutilised double track railway.

 

[Kentish Paul]

Typically, it is up trains that are looped, and they are timed to wait for 5 minutes which is clearly added to their overall journey time as they also stop at the station just beforehand.

Not in my travels on the line. My down trains are looped more often than not.

 

[MarkyT]

Not in my travels on the line. My down trains are looped more often than not.

Apologies. You are right and I was wrong! I based my assumption on an RTT search for today, completely forgetting it was a Sunday and there's a special service operating with trains going no further east than Basingstoke. Searching again for other dates I see the majority of weekday looped trains are down services as you say, and the loop stop is timed for four minutes. The same applies on a 'normal' Sunday.

 

[mr_jrt]

Another constraint for Penryn style is platform length and whether space is available for the extra long construction. Even with only very short trains the combined length of the sequential platforms adds up to 238m at Penryn itself. On many branch lines, that kind of length may not be available between other obstacles at either end like bridges or level crossings.

Handy for when the line moves to 12-car trains though

 

[Searchlight]

Just to back up what you've said.

Thanks for the correction. These longer type dynamic loops are really what I had in mind.......Not neccessariliy 3 miles though ! If there was a single platform station between two of these loops, that would be equilvalent to what I proposed.


Say you were building a NEW railway between two traffic centres, via several small towns. Would it be cheaper to build a conventional (British) double track line with two platform stations, footbridges, etc.
OR would it be more cost effective to build a single line with single platform stations and frequent loops? As indeed many countries do.....That is the real question. As has been pointed out, the Penryn set-up is only really suitable for branch lines with very short trains.

For the sort of line I have mentioned, trains would be four five or six cars long........

 

[MarkyT]

Thanks for the correction. These longer type dynamic loops are really what I had in mind.......Not neccessariliy 3 miles though ! If there was a single platform station between two of these loops, that would be equilvalent to what I proposed.

Effectively, that is how the Salisbury - Exeter line works today, with some stations on the single line and others within the fairly long sections of double track.

Most normal passes occur on the longish double track sections, and these occur as regularly as clockwork throughout the day with the long loop positions tuned to the hourly clockface service pattern. Whilst also a regular passing point, as noted earlier Tisbury Loop is a short slow pull-in just London side of the single platform station. Down trains are usually looped and have to wait a few minutes there before drawing into the station once the up train has left. One of the early options for Axminster was a dynamic loop between stations to avoid any station rebuilding. Operators much preferred a long loop at Axminster station however. That neatly places the next regular passing position towards London midway along the next double track section between Yeovil and Templecombe. The trains actually pass just to the west of Sherborne station normally. A big advantage of the dynamic system in addition to the absence of a journey time penalty is there is a degree of resilience built in so any minor delay in one direction is not passed on to the other automatically. There are some additional loops west of Yeovil remaining at Chard Junction (not a station) and Honiton, but these are not used by the regular clock face pattern, only by a few peak variations and irregular empty stock and departmental movements. Otherwise, they can be used to keep late trains moving forward. I remember returning from London one Sunday evening. We were detained at Yeovil due to a late running up following a signal failure. After leaving, we were pulled in at Chard Junction where we were passed by the next up estimated running only a few minutes late. Not passing anything at Axminster, next we were pulled in at Honiton where it was announced we would wait for the next London train to pass, which it did on time. When we arrived at Exeter we were some 40 minutes late but the set turned straight around and departed on the next run back to Salisbury only a couple of minutes late. We were the sacrificial train that gathered more delay yet was kept moving forward enabling the rest of the service pattern to recover very quickly.

Say you were building a NEW railway between two traffic centres, via several small towns. Would it be cheaper to build a conventional (British) double track line with two platform stations, footbridges, etc.
OR would it be more cost effective to build a single line with single platform stations and frequent loops? As indeed many countries do.....That is the real question. As has been pointed out, the Penryn set-up is only really suitable for branch lines with very short trains.

For the sort of line I have mentioned, trains would be four five or six cars long........

You are describing the new Borders line which is indeed mostly single track with some dynamic loops where trains are planned to pass on the move regularly. The northernmost of those loops has no stations within it, but I think that was because there was no viable site for one in the vicinity rather than any particular desire to save costs by not putting stations on the double track.

 

[fegguk]

You are describing the new Borders line which is indeed mostly single track with some dynamic loops where trains are planned to pass on the move regularly. The northernmost of those loops has no stations within it, but I think that was because there was no viable site for one in the vicinity rather than any particular desire to save costs by not putting stations on the double track.

This is not quite correct the northern loop has a 2 platform station Shawfair. The middle loop has no stations, though Gorebridge is very near is northern end, the Southern loop has a 2 platform station at Stow. All the other intermediate station have one platform only. The Tweedbank terminus station has a 2 platform island configuration. The loops are variable in length the longest several miles long.

 

[Searchlight]

Well they reckon wrong. On any given single track section, capacity cannot be more than 50% of a double track section with the same signalling headway.

Capacity on long freight railways across the prairies is a rather different concept to capacity on a predominantly passenger railway with lots of stations and junctions.

I was specifically refering to CTC single line operations. If you look up
CTC operation, you will see figures of 70% 75% quoted regarding track capacity. This can be easily explained, because we think of single line as just that, whereas in the USA, Canada, Australia etc. They think of long dynamic loops, full track circuiting and local interlockings. Often they will initially build a "single track" line, and, only upgrade to double track if the traffic demands it. Often it doesent. Loops in the USA are often 2 miles long.......With some freight trains 1 mile long, such loops are necessary.

I contend we never have really given CTC a fair trial in the UK over a significant distance.....It is an "alien" concept here, with double track lines being the norm. This is why people assume single lines have 50% capacity. British single lines tend to be under engineered regarding passing loops and lack CTC.

The Metropolitan Railway DID experiment with CTC on the Stanmore Branch in 1932, albeit a double track line.

 

[najaB]

If you look up CTC operation, you will see figures of 70% 75% quoted regarding track capacity.

The point Bald Rick was making is that a double-track line will always have at least double the capacity of a single-track line for traffic in opposite directions. You can only run trains in one direction per track at any given time (well, you can try to run in both but it seriously reduces the resale value of the rolling stock). If you have two tracks you have twice the capacity.

 

[OxtedL]

I contend we never have really given CTC a fair trial in the UK over a significant distance.....It is an "alien" concept here, with double track lines being the norm. This is why people assume single lines have 50% capacity. British single lines tend to be under engineered regarding passing loops and lack CTC.

I think you are wrong, because by CTC I think you essentially mean a long section of track(s) dispatched by one person (who would be the signaller, in the UK).

Examples of single track with passing loops with some kind of central signal control (rather than, say, a separate signal box at each passing point) include the Cambrian line (on ERTMS), the West Highland and Far North Lines in Scotland (on RETB) and in a less remote area the Uckfield line (conventional signals controlled from Oxted).

There are very few remaining locations in the UK with double track that could be legitimately replaced by single track and passing loops without big capacity problems, even with improved central control. The only example I can think of is the line between Giggleswick-Carnforth, although there are possibly others.

(If you need any of those terms or examples explaining then say and someone can help.)

 

[Philip Phlopp]

I contend we never have really given CTC a fair trial in the UK over a significant distance....

There's a reason for that - it's because it's crap, quite simply. It's no use trying to explain away that fact, no matter how you say it, CTC is a crap concept and not worth wasting either the railway's time and money, or this forum's bandwidth even thinking about any further.

 

[Joseph_Locke]

At a detailed performance level it can be argued (with plenty of justification) that a busy single line with a generally even and regular flow in each direction has a slightly lower capacity than 50% of a twin track layout as the time of operation for S&C, overlap clearance times and time spent waiting for opposite direction moves to clear all eat into performance.

It also has two point ends extra, which are additional points (!) of failure, leading to further unavailability.

CTC is a traffic management system of sorts, not perpetual motion.

 

[MarkyT]

I was specifically refering to CTC single line operations. If you look up CTC operation, you will see figures of 70% 75% quoted regarding track capacity.

America again, where most frequent passenger operations were curtailed in the 1960s. There were many cases of singling of former multitrack routes after that whilst railroads continued to carry just as much if not more tonnage than they did before. Equally many rural routes had only carried very infrequent passenger traffic anyway even in their heyday. Without the pressures of those pesky public passenger schedules, railroads gained much more flexibility to run their remaining freight traffic when it was convenient for them. The traditional hub and spoke networks typically had long connecting windows between trains and customer expectations of many days in transit, so precise arrival times in yards were much less critical than any passenger operation. Sparse infrastructure and long waits in country sidings are common in much of the US. The long stretches of single track, albeit sometimes with impressive dynamic loops in places are probably the main challenge to on-time performance for most of Amtrack's long distance trains. If you miss your slot you may have to wait for an opposite direction freight to clear a 20-mile section. Half an hour delay if you're lucky; much more perhaps if it's hilly. Traffic growth and a generally tighter more logistical approach to freight shipments has put pressure on railroads to operate more precisely and additional capacity to help with that has been installed where possible, sometime redoubling former multi-tracks, sometimes adding entirely new sections. Sometimes innovative operating ideas have been adopted such as tidal flows. These are often more feasible with freight where you could run in one direction for so many hours then switch to the other direction for another long period. That approach is rarely possible with passenger operations, unless the passenger trains are particularly infrequent. Elsewhere alternative parallel single lines have been dedicated one to each direction. Opposing traffic can still run on each route if necessary but it's given lower priority leading to long waits in loops. Competing railroads have even been known to use running powers to cooperate in sharing parallel routes in this way leading to greater capacity and improved transit time for them both.

I contend we never have really given CTC a fair trial in the UK over a significant distance.....It is an "alien" concept here, with double track lines being the norm. This is why people assume single lines have 50% capacity. British single lines tend to be under engineered regarding passing loops and lack CTC.The Metropolitan Railway DID experiment with CTC on the Stanmore Branch in 1932, albeit a double track line.

Early American CTC is the ancestor and prototype for all modern signalling ever since all over the World, and note that in the 1930s and 1940s US railroads installed it over many thousands of miles of multiple track main line years before the postwar passenger cutbacks and simplifications. Once the railroads demonstrated the system's possibilities, all nations started to adopt the centralised control ideas very quickly, UK included, but mostly for our multitrack main lines where reduction of the number of running lines was not possible widely, but facilities at junctions and stations were reduced based on the greater throughput efficiency claimed for modern signalling. Many modern single lead junctions attest to that. The UK has always been slow, comparatively, to roll out centralised control on rural and secondary routes however, often finding it difficult to justify the capital expense. The Salisbury to Exeter single line (west of Tisbury) with it's dynamic and other loops is now finally (since 2012) controlled from Basingstoke ASC (Area Signalling Centre) instead of a number of local signal boxes. The lack of centralised control has not prevented the route performing well as a largely single track long distance corridor since simplification in the late 1960s (it was formerly double track all the way). Since 2009 the line has hosted an hourly clockface timetable all the way to Exeter, but there's no doubt the simplicity of the operation, with an hourly repeating timetable and no other regular traffic with different stopping patterns or journey times makes this possible.

Wherever they are in the World and however centralised their control, railways tend to evolve to the best form layout and number of tracks needed to handle the types and mix of traffic the railway or other governing agency decides it will operate economically. In UK, I doubt there are many remaining places further reductions of existing double tracks would be possible today in a time of traffic growth and increasing concern over performance. That's not to say some short stretches couldn't be justified where they offer some other important benefit such as junction or station simplification, speed increases in tightly constrained topographies or to overcome some structural or clearance concern in some otherwise very expensive to rebuild asset. On the Scottish Borders Railway I think a view has been taken to include only as much double track as absolutely necessary for the comparatively self-contained initial service. Whilst more might be nice, it simply can't be justified, so that is a very good example of the UK using your approach to minimise costs by eliminating surplus (new) double track whilst still being able to offer a modern fast service performance with passenger trains passing at speed in fairly long dynamic loops. Whilst the service concept remains as it is the infrastructure configuration is perfectly adequate.

 

[MarkRedon]

I contend we never have really given CTC a fair trial in the UK over a significant distance.....It is an "alien" concept here, with double track lines being the norm. This is why people assume single lines have 50% capacity. British single lines tend to be under engineered regarding passing loops and lack CTC.

There are very few remaining locations in the UK with double track that could be legitimately replaced by single track and passing loops without big capacity problems, even with improved central control. The only example I can think of is the line between Giggleswick-Carnforth, although there are possibly others.

CTC might perhaps have a role on what are fundamentally freight-only lines where competitive pressures on pricing favour lower costs and accept poor transit times. That is almost everywhere the case in the USA and almost never the case in Britain. The logic of increasingly centralised signalling and, in due course, ERTMS in the UK will permit some of the benefits of CTC to be realised here in any event.

Can someone please inform us what the cost savings associated with singling actually are? I assume they are substantially less than 50%.

On a very parochial note. The (Settle Junction to) Giggleswick to Carnforth line is one of my very favourite British railways. It was intended to single the line in the early 1980s at much the same time as the intention was to close completely the Settle Junction to Carlisle line. I understand singling did not proceed because the expectation then developed that the line to Carnforth would also completely close. When subsequently the battle to retain the S&C was won, a very slow single line junction at Settle Junction to the Carnforth and Morecambe line was installed so as to reduce maintenance costs. The line as a whole is in reasonably good fettle, not least because it is very lightly trafficked. I sincerely hope that not even this rather rural railway needs to be singled in its entirety.

 

[Searchlight]

America again, where most frequent passenger operations were curtailed in the 1960s. There were many cases of singling of former multitrack routes after that whilst railroads continued to carry just as much if not more tonnage than they did before. Equally many rural routes had only carried very infrequent passenger traffic anyway even in their heyday. Without the pressures of those pesky public passenger schedules, railroads gained much more flexibility to run their remaining freight traffic when it was convenient for them. The traditional hub and spoke networks typically had long connecting windows between trains and customer expectations of many days in transit, so precise arrival times in yards were much less critical than any passenger operation. Sparse infrastructure and long waits in country sidings are common in much of the US. The long stretches of single track, albeit sometimes with impressive dynamic loops in places are probably the main challenge to on-time performance for most of Amtrack's long distance trains. If you miss your slot you may have to wait for an opposite direction freight to clear a 20-mile section. Half an hour delay if you're lucky; much more perhaps if it's hilly. Traffic growth and a generally tighter more logistical approach to freight shipments has put pressure on railroads to operate more precisely and additional capacity to help with that has been installed where possible, sometime redoubling former multi-tracks, sometimes adding entirely new sections. Sometimes innovative operating ideas have been adopted such as tidal flows. These are often more feasible with freight where you could run in one direction for so many hours then switch to the other direction for another long period. That approach is rarely possible with passenger operations, unless the passenger trains are particularly infrequent. Elsewhere alternative parallel single lines have been dedicated one to each direction. Opposing traffic can still run on each route if necessary but it's given lower priority leading to long waits in loops. Competing railroads have even been known to use running powers to cooperate in sharing parallel routes in this way leading to greater capacity and improved transit time for them both.



Early American CTC is the ancestor and prototype for all modern signalling ever since all over the World, and note that in the 1930s and 1940s US railroads installed it over many thousands of miles of multiple track main line years before the postwar passenger cutbacks and simplifications. Once the railroads demonstrated the system's possibilities, all nations started to adopt the centralised control ideas very quickly, UK included, but mostly for our multitrack main lines where reduction of the number of running lines was not possible widely, but facilities at junctions and stations were reduced based on the greater throughput efficiency claimed for modern signalling. Many modern single lead junctions attest to that. The UK has always been slow, comparatively, to roll out centralised control on rural and secondary routes however, often finding it difficult to justify the capital expense. The Salisbury to Exeter single line (west of Tisbury) with it's dynamic and other loops is now finally (since 2012) controlled from Basingstoke ASC (Area Signalling Centre) instead of a number of local signal boxes. The lack of centralised control has not prevented the route performing well as a largely single track long distance corridor since simplification the late 1960s (it was formerly double track all the way). Since 2009 the line has hosted an hourly clockface timetable all the way to Exeter, but there's no doubt the simplicity of the operation, with an hourly repeating timetable and no other regular traffic with different stopping patterns or journey times makes this possible.

Wherever they are in the World and however centralised their control, railways tend to evolve to the best form layout and number of tracks needed to handle the types and mix of traffic the railway or other governing agency decides it will operate economically. In UK, I doubt there are many remaining places further reductions of existing double tracks would be possible today in a time of traffic growth and increasing concern over performance. That's not to say some short stretches couldn't be justified where they offer some other important benefit such as junction or station simplification, speed increases in tightly constrained topographies or to overcome some structural or clearance concern in some otherwise very expensive to rebuild asset. On the Scottish Borders Railway I think a view has been taken to include only as much double track as absolutely necessary for the comparatively self-contained initial service. Whilst more might be nice, it simply can't be justified, so that is a very good example of the UK using your approach to minimise costs by eliminating surplus (new) double track whilst still being able to offer a modern fast service performance with passenger trains passing at speed in fairly long dynamic loops. Whilst the service concept remains as it is the infrastructure configuration is perfectly adequate.

Thank you for your detailed and balanced reply. You put across the issues very well. I don t envison many more singling schemes in Britain, but new schemes like the Borders Line, would be candidates for evaluation of the pros and cons of single lines or partial singling.

ERTMS is obviously the way forward for future signalling. Existing single lines should be looked at, with a view to improving capacity. (Not neccessarily by doubling)

 

[OxtedL]

I don't envison many more singling schemes in Britain, but new schemes like the Borders Line, would be candidates for evaluation of the pros and cons of single lines or partial singling.

ERTMS is obviously the way forward for future signalling. Existing single lines should be looked at, with a view to improving capacity. (Not neccessarily by doubling)

Recent schemes like the Borders Railway and Ebbw Vale Line have used single track, as you would promote, so it is clear that schemes are already evaluated in the way you envisage.

What is found is that if you ever need to run more trains in the future, you have to double more track! When a single track is being timetabled to be used at full capacity there are no other choices, it is extremely rare that you can fit more trains on just with snazzier signalling.

Other schemes like the Airdrie-Bathgate reopening used double throughout, as it was possible and better suited the frequent nature of trains on the line.

I think the issues you are describing are already well understood by those in charge of designing new rail lines, and you are not proposing anything new.