What are "dynamic loops" on single line branches?

[imt]

I am trying to understand the process of increasing the frequency of trains on a single bi-directional line and hoping some of the experts would have the time to help me. Such a line would be controlled by various means - Track Circuit Block, Token Block, Tokenless Block, Radio Token Block (and I am sure the experts could name more). The objective being that once a train had entered the single line another could not enter from the other end until the first had (completely) passed it by.  

 

One way of providing increased frequency is to put in a passing loop - which may or may not be at a station: effectively splitting one single line section into two, presumably at the cost of another signal box and signaller. I suppose (unless it was a station that HAD to be sited where it was) the loop would be put as near half-way as possible and then (discounting delays for stopping before proceeding into the loop etc. "BR Regs. for Train Signalling on Single Lines" BR29960/15) you can have double the service. Some of you kindly enlightened me on many aspects of Tokenless Block in this thread

 

http://www.rmweb.co.uk/community/index.php?/topic/113085-information-please-on-scottish-tokenless-block-working/

 

So I am moving on from there.

 

I keep seeing in these pages and elsewhere (e.g. plans for the Inverness - Aberdeen line, but there are others) the words "Dynamic Loop" being used.  This doesn't seem very descriptive - unless they get up and move? (Sorry for the silly joke!).

 

As I understand it, a "dynamic loop" is simply a length of double track (as long as you can make it weighed on cost benefit I expect).  I presume (depending on length) that maybe two or more trains could be moving in each direction, and possibly the key aspect is that there is no delay for a train approaching the loop in entering the loop - since the approaching train could be several miles away rather than impatiently waiting at the loop exit signal.  However the additional cost seems to be that there must now be 2 signal boxes and signallers, one at each end of the dynamic loop.  Now I am sure modern electronic system would dispense with signal boxes and have a signal control centre, but how would the late 60s/70s have dealt with the problem? I presume MAS would be involved here to allow for the extra density of traffic?

[The Stationmaster]

A dynamic loop is designed in such a way that, on a single line, crossing trains should pass each other without needing to stop.  Thus the loop has to be of sufficient length, and with relatively high speed (in relation to linespeed) turnouts plus the correct arrangement of signalling and overlaps in order to facilitate that.

 

Dynamic loops can also be used to provide looping off running lines.  For example when I developed the infrastructure and trainplans for imported coal to Didcot power station I incorporated dynamic loops by having the former Relief Lines between Wantage Road and Challow reinstated - the timetabling part of the plan meant that a loaded coal train could leave Swindon and enter the Up 'loop' (classified once more as a Relief Line) at Challow and continue at its maximum speed to wantage Road and come out behind an overtaking HST - assuming it had left Swindon at the correct margin ahead of the HST.  Somewhere I still have the rather detailed graph I drew to work out the precise margin that would allow that.

 

Regrettably the dynamic loop enhancement I came up with (as a consultant in that instance for teh restoration of passenger trains on the Western Valley (to Ebbw Vale) did not proceed in teh way I recommended in order to save money.  However if extra services are added to more destinations it would make a lot of sense to do it as I planned it.

[imt]

A dynamic loop is designed in such a way that, on a single line, crossing trains should pass each other without needing to stop.  

 

 

So in fact a Dynamic Loop would be as well NOT to have a station in it since that could (potentially) foul up the timings.  Obviously this is down to good timetable planning, but avoidance would be better? 

[brigo]

So in fact a Dynamic Loop would be as well NOT to have a station in it since that could (potentially) foul up the timings.  Obviously this is down to good timetable planning, but avoidance would be better? 

 

A dynamic loop can include a station. An example is on the West Midlands Cross City line which runs south from Lichfield through New Street to Redditch. After the junction at Barnt Green the line was single through Alvechurch to Redditch. This restricted trains to 2 per hour on this section. Recently a dynamic loop was built from just north of Alvechuch and then run south for about 2 miles. Alvechurch station gained another platform, overbridge and lifts as required these days. The result is that train frequency has been raised to 3 per hour.

 

Brian

[PaulRhB]

A dynamic loop with a station in the middle means a train can leave on time and get up to speed without a direct knock on from an opposing train running a few minutes adrift. All it means is the trains cross nearer the end of the loop rather than at the station. It does rely on the meets being timed at the station and not at the junctions though or all you've really got is a short length of double track line!

[imt]

A dynamic loop can include a station. An example is on the West Midlands Cross City line which runs south from Lichfield through New Street to Redditch. After the junction at Barnt Green the line was single through Alvechurch to Redditch. This restricted trains to 2 per hour on this section. Recently a dynamic loop was built from just north of Alvechuch and then run south for about 2 miles. Alvechurch station gained another platform, overbridge and lifts as required these days. The result is that train frequency has been raised to 3 per hour.

 

Brian

 

Thanks for that, its interesting to see that they are (still?) being implemented and the effect on traffic that they have.  The original line from Inverness to Dingwall was double, then singled, then doubled again (repeat as nauseam) between Clachnaharry and Clunes which included the now closed Lentran station.  Its currently single and a horrible bottleneck especially with the revival of rail in this part of Scotland: to be another of those much discussed "dynamic loops" now.  The line between Blair and Dalwhinnie stations (something like 25 miles) is dual whilst much around it is single.  I note many of those pressing for improvements ask for doubling of lines and where that is not possible "dynamic loops" - so I am still a little confused as to which is what and why.

[imt]

A dynamic loop with a station in the middle means a train can leave on time and get up to speed without a direct knock on from an opposing train running a few minutes adrift. All it means is the trains cross nearer the end of the loop rather than at the station. It does rely on the meets being timed at the station and not at the junctions though or all you've really got is a short length of double track line!

 

Thanks for that insight.  As far as I can see much of this is about sensible timetabling as opposed to just some double track.  Getting the paths right and their crossings scheduled intelligently is important.  One of the problems on the Perth-Inverness is extended waiting in stations for passing trains - what you are suggesting above would have quite an impact since there is extended passing time available, so if one is late/early it can to some extent be absorbed by the stretch of double line.

[Zomboid]

Axminster is another example of a dynamic loop. As is Templecombe to Yeovil In.

I think the emphasis is more on both trains being able to go about their business without impacting each other, rather than without stopping.

[Dunsignalling]

 

I am trying to understand the process of increasing the frequency of trains on a single bi-directional line and hoping some of the experts would have the time to help me. Such a line would be controlled by various means - Track Circuit Block, Token Block, Tokenless Block, Radio Token Block (and I am sure the experts could name more). The objective being that once a train had entered the single line another could not enter from the other end until the first had (completely) passed it by.  

 

One way of providing increased frequency is to put in a passing loop - which may or may not be at a station: effectively splitting one single line section into two, presumably at the cost of another signal box and signaller. I suppose (unless it was a station that HAD to be sited where it was) the loop would be put as near half-way as possible and then (discounting delays for stopping before proceeding into the loop etc. "BR Regs. for Train Signalling on Single Lines" BR29960/15) you can have double the service. Some of you kindly enlightened me on many aspects of Tokenless Block in this thread

 

http://www.rmweb.co.uk/community/index.php?/topic/113085-information-please-on-scottish-tokenless-block-working/

 

So I am moving on from there.

 

I keep seeing in these pages and elsewhere (e.g. plans for the Inverness - Aberdeen line, but there are others) the words "Dynamic Loop" being used.  This doesn't seem very descriptive - unless they get up and move? (Sorry for the silly joke!).

 

As I understand it, a "dynamic loop" is simply a length of double track (as long as you can make it weighed on cost benefit I expect).  I presume (depending on length) that maybe two or more trains could be moving in each direction, and possibly the key aspect is that there is no delay for a train approaching the loop in entering the loop - since the approaching train could be several miles away rather than impatiently waiting at the loop exit signal.  However the additional cost seems to be that there must now be 2 signal boxes and signallers, one at each end of the dynamic loop.  Now I am sure modern electronic system would dispense with signal boxes and have a signal control centre, but how would the late 60s/70s have dealt with the problem? I presume MAS would be involved here to allow for the extra density of traffic?

 

Back in the 60s/70s the rail imperative was generally based on retrenchment / singling / closure rather than expansion or service enhancement, so the situation wouldn't have arisen. However, so long as the installation had power points and colour light signalling, there is no reason why a dynamic loop could not be operated from a single box.

 

The number and spacing of signals has to exceed the maximum expected demand for pathways or timetabling becomes so hedged about with conditions and that any gain quickly disappears.

 

The one with which I am familiar, at Axminster, can theoretically hold three trains in each side but I can't envisage any need/ability arising to feed that many into and out of it via single line approaches unless the adjacent conventional loops at Chard Jn. and Honiton were both out of action

 

In practice, two in one direction and one in the other is the usual maximum, though two each way may have occurred once or twice.

 

Both roads are signalled bi-directionally, too, but beyond formerly catering for one terminating train a day from Exeter (which no longer goes back!), it doesn't do much other than saving passengers boarding early morning up trains having to use the footbridge when there are no down trains about.

 

It also doesn't provide the full benefit it was designed to offer as some bright spark decided to save a few grand by downgrading the points from the intended 70mph rated type to 50mph ones.

 

John

 

Sorry for all the edits. I was in the middle of posting when the site went down.

Edited June 24, 2017 by Dunsignalling

[imt]

Thanks for your response - I end up having to edit things too so I wouldn't worry.  Your comments are in italics below:

 

 

Back in the 60s/70s the rail imperative was generally based on retrenchment / singling / closure rather than expansion or service enhancement, so the situation wouldn't have arisen. 

 

However, so long as the installation had power points and colour light signalling, there is no reason why a dynamic loop could not be operated from a single box.

 

In my earlier thread I explained about my hypothetical railway branch - so I was really thinking about the late 60s/70s and the consequences of increased rathere than collapsing traffic (my own quiet never-never land!).  Useful to see that remote control of points etc would have been OK.  The Stationmaster had already reminded me that higher speed junction points would also help by reducing the need to slow trains approaching them. 

 

The number and spacing of signals has to exceed the maximum expected demand for pathways or timetabling becomes hedged about with conditions and any gain quickly disappears.

 

The one with which I am familiar, at Axminster, can theoretically hold three trains in each side but any need/ability to feed that many into and out of it via single line approaches seems extremely unlikely unless the adjacent conventional loops at Chard Jn. and Honiton were both out of action

 

This is how long is a piece of string of course, but I was looking to get some idea of how long such a "dynamic loop" would need to be to hold say 2 trains and keep them moving - 3 miles?

 

In practice, two in one direction and one in the other is the usual maximum, though two each way may have occurred once or twice.

 

Both roads are signalled bi-directionally, too, but beyond one terminating train a day from Exeter, it doesn't do much other than saving passengers boarding early morning up trains having to use the footbridge when there are no down trains about.

 

I had thought about bi-directional working, but thought that such signalling would be a very expensive luxury.

[Dunsignalling]

 

Thanks for your response - I end up having to edit things too so I wouldn't worry.  Your comments are in italics below:

 

 

Back in the 60s/70s the rail imperative was generally based on retrenchment / singling / closure rather than expansion or service enhancement, so the situation wouldn't have arisen. 

 

However, so long as the installation had power points and colour light signalling, there is no reason why a dynamic loop could not be operated from a single box.

 

In my earlier thread I explained about my hypothetical railway branch - so I was really thinking about the late 60s/70s and the consequences of increased rathere than collapsing traffic (my own quiet never-never land!).  Useful to see that remote control of points etc would have been OK.  The Stationmaster had already reminded me that higher speed junction points would also help by reducing the need to slow trains approaching them. 

 

The number and spacing of signals has to exceed the maximum expected demand for pathways or timetabling becomes hedged about with conditions and any gain quickly disappears.

 

The one with which I am familiar, at Axminster, can theoretically hold three trains in each side but any need/ability to feed that many into and out of it via single line approaches seems extremely unlikely unless the adjacent conventional loops at Chard Jn. and Honiton were both out of action

 

This is how long is a piece of string of course, but I was looking to get some idea of how long such a "dynamic loop" would need to be to hold say 2 trains and keep them moving - 3 miles?

 

In practice, two in one direction and one in the other is the usual maximum, though two each way may have occurred once or twice.

 

Both roads are signalled bi-directionally, too, but beyond one terminating train a day from Exeter, it doesn't do much other than saving passengers boarding early morning up trains having to use the footbridge when there are no down trains about.

 

I had thought about bi-directional working, but thought that such signalling would be a very expensive luxury.

 

I'm not certain of the exact length of the Axminster one; I think it's a bit over three and a half miles.

 

The bi-di will come into its own if/when the proposed Devon Metro service from Exeter (or possibly Barnstaple) to Axminster gets slotted in between the hourly Waterloo traffic.

 

Up to now, the only time the capacity of the loop really helps is in service recovery or handling diversions off the GW line. The main benefit is keeping things moving and preventing smallish knock-on delays building up.

 

John

Edited June 24, 2017 by Dunsignalling

[The Johnster]

I'm familiar with the ones on the Ebbw Vale line that weren't built to Mike's specifications.  The system clearly relies on both trains entering the dynamic loop at or very close to the times specified in the WTT; once in, it is of no importance in terms of causing delay to the other train what time they leave...  it works on the Ebbw Vale service because, whilst occupation is almost at saturation, speeds are relatively low and the trains spend some time on the loops, but as Mike says expansion will be needed if the service is to be increased, which I believe is on the cards, and his design may well be needed after all.

 

The original Ebbw Vale passenger scheme had all the hallmarks of a cut price job which was a laudable attempt to find a use for a redundant freight branch after Ebbw Vales steelworks closed, and has caught it's operators our with it's success and popularity.  I would look for speed increases on some section of it to increase capacity and make journey times to Cardiff more attractive in competition with the bus, which will place more pressure on the existing loops.  There seem no reason beyond lack of cash that the line could not be reinstated as double track throughout until you get to some new bridges in Ebbw Vale itself; trackbed is still in situ.

[Dungrange]

This is how long is a piece of string of course, but I was looking to get some idea of how long such a "dynamic loop" would need to be to hold say 2 trains and keep them moving - 3 miles?

 

Instead of thinking about how long a dynamic loop might be in miles, think about how much resilience you want and how much you are willing to pay for that.

 

Two trains should ideally be timetabled to enter the dynamic loop, one from each direction, at around the same time.    If both are running according to the timetable then they will pass within the extents of the dynamic loop and continue on their journey without any delay.  If the line speed is 60 mph, then a train travelling at the line speed will cover one mile in one minute.  Therefore, if a dynamic loop is just over three miles long, then it will take a train three minutes to travel from the start of the loop to the end at 60 mph.  What this means is that unless there is a scheduled station stop within the extents of the dynamic loop, a service will still be delayed if the train running in the opposite direction is running more than three minutes late.  If you want more resilience, then you need to increase the length of the dynamic loops.  If you want a higher line speed and the same level of resilience, then you need to increase the length of the dynamic loops.  Increasing the length of the loops will increase the cost and the optimum length will be a balance between cost and resilience.  I don't think that there is any specific threshold at which a dynamic loop becomes a short length of double track.  Indeed, I'm sure that I've seen the section of double track between Dalwhinnie and Blair Atholl referred to as dynamic loops (which would clearly allow for maybe 20 minutes of delay before the train travelling in the opposite direction is affected.

[The Stationmaster]

 

 

 

The number and spacing of signals has to exceed the maximum expected demand for pathways or timetabling becomes hedged about with conditions and any gain quickly disappears.

 

The one with which I am familiar, at Axminster, can theoretically hold three trains in each side but any need/ability to feed that many into and out of it via single line approaches seems extremely unlikely unless the adjacent conventional loops at Chard Jn. and Honiton were both out of action

 

This is how long is a piece of string of course, but I was looking to get some idea of how long such a "dynamic loop" would need to be to hold say 2 trains and keep them moving - 3 miles?

 

 

 

 

The length of a dynamic loop depends on a number of factors of which the most important is the end to end time of what you might term 'a normal service train' and then the potential frequency.  Now what this means is that in many respects if you plan a dynamic loop to very closely match particular timetable requirements and particular types of trains it will no longer work properly if, say, train/linespeeds increase and/or frequency increases.

 

This is where the Ebbw Vales situation comes home to roost as what was actually installed was done down to a price to match a particular pattern of train service although it does have limited flexibility.  If you need to run more trains (in its case to different destinations) then you will need to go back and lengthen the loop and alter the signalling in order to both increase loop capacity and still maintain a dynamic capability.  None of it is exactly difficult to work out if you know how to do it which means understanding a whole variety of inputs and how they relate to each other.

[28XX]

The Cotswold line has two dynamic loops. Ascot Under Wychwood to Charlbury and Moreton in Marsh to a point West of Evesham. Works well.

 

Edit: Had changed to has.

Edited June 25, 2017 by 28XX

[NittenDormer]

I am familiar with the two 'dynamic loops' on the Borders railway - one from Gorebridge up Borthwick Bank, and the other around Stow (including a station, although not every service stops there). They allow north/south services to pass at full speed, although there has been criticism. I know the Edinburgh bound services sometimes have to stop outside Gorebridge to wait for the Tweedbank train to enter the double track section. Also, I regularly hear trains passing non-stop (when they should be stopping), presumably because they need to make up time. The single track sections seem to cause problems getting the timetable back on track after a delay.

[imt]

I am familiar with the two 'dynamic loops' on the Borders railway - one from Gorebridge up Borthwick Bank, and the other around Stow (including a station, although not every service stops there). They allow north/south services to pass at full speed, although there has been criticism. I know the Edinburgh bound services sometimes have to stop outside Gorebridge to wait for the Tweedbank train to enter the double track section. Also, I regularly hear trains passing non-stop (when they should be stopping), presumably because they need to make up time. The single track sections seem to cause problems getting the timetable back on track after a delay.

 

Thank you, I'll take a closer look at that line because there seems to be quite a bit of material - presumably because its new.  You have hit the nail on the head with single lines - any "significant" delay (and it begins to look like that delay time will vary from line to line, section to section) will (gradually?) create chaos.  The Swiss, who seem to have a lot of single line sections, seem to manage OK (frequent visitor) but there are times when you wait for passing trains.  They also don't aim for 100mph trains, just a timetable that works and interlinks with all other forms of transport.

[Edwin_m]

As suggested above the length of a dynamic loop should really be measured in travel time rather than distance (subject to some complication in relation to signal spacing and overlaps to ensure both trains can run through on greens).  This means that if there is a station on the loop where many or all trains stop, the loop itself can be shorter and cheaper for the same performacne benefit.  However in cost terms this is counterbalanced by the need to provide a second platform and possibly an expensive fully-accessible footbridge at the station, which may not be possible if land isn't available. 

[Joseph_Pestell]

Going abroad, when the main line from Paris - Reims was redesignated as being via Epernay (to reduce electrification costs) instead of via Ferte-Milon, dynamic loops were put in between Epernay and Reims. With bi-directional signalling, it enabled a fairly frequent service where fast mainline trains (Paris - Charleville - Luxembourg) could overtake stopping locals and freight trains.