Freight services being updated

[iands]

It relies on 100% availability of the radio system (temporary black-outs tolerated). For the standard system this is GSM-R, but outside the EU there is a hybrid system using TETRA rather than GSM-R. It is in one of the -stans. I cannot remember of it is Kazakhstan or Tajikistan.

To clarify, "availability" requirement of the GSMR is "better than 99%". Coverage probability (of the railway) of GSMR is; 95% based on a coverage level of -98dBm for voice and non-safety critical data, and 95% based on a coverage level of -95dBm on lines with ETCS level 2/3 for speeds lower than 220km/h.

 

To slightly confuse matters, GSMR by itself (e.g. the Base Transceiver Stations) cannot support ETCS as the BTSs don't have the required bandwidth available for the ETCS data transmissions. Additional kit, which effectively sits along side the BTSs but connected to the same antenna infrastructure, will be required to support the ETCS data transmissions.

 

As an aside, I understand Finland has abandoned (or is abandoning) GSMR in favour of an alternative system, but I'm not sure if it is TETRA or another system.

 

Regards, Ian.

[Supaned]

I'd also add that as good as GSM-R is (and a million times better than NRN) , it is still by no means perfect - loss of network still occurring fairly frequently in some locations - personally, I'd want any sort of signalling kit that uses that technology to be more reliable.

[Rugd1022]

The dead spots in the GSM-R network can be a right pain in the Aris', three which spring to mind on my route card are Leicester, nearby Kirby Muxloe on the Burton branch and the area around Willesden Euro Terminal. If I hit the 'SG' button on the Willesden Relief lines alongside the terminal it puts me through to Acton Canal Wharf 'box instead of Wembley Mainline. I know it's possible to sidestep this by finding and scrolling through the list of numbers in the display, but it's actually quicker to get down on the ballast and use the old SPT...!

Edited January 5, 2018 by Rugd1022

[iands]

I'd also add that as good as GSM-R is (and a million times better than NRN) , it is still by no means perfect - loss of network still occurring fairly frequently in some locations - personally, I'd want any sort of signalling kit that uses that technology to be more reliable.

No radio system is "perfect", they will all suffer from the effects of some sort of interference from time to time (and in some cases the "interference" can cause a temporary "black spot"), as well as some other external influence causing a more seemingly permenent black spot. These usually occur because something has changed since the original radio survey and build was undertaken - like a new building(s) sprouting up, lineside vegitation growing unchecked for several years, as well as a general degradation over time of the radio antenna and feeder cable system (both lineside kit and train on-board kit) - all of these conspire to creating a black spot somewhere on the network at any given time.

 

Regards, Ian.

[iands]

The dead spots in the GSM-R network can be a right pain in the Aris', three which spring to mind on my route card are Leicester, nearby Kirby Muxloe on the Burton branch and the area around Willesden Euro Terminal. If I hit the 'SG' button on the Willesden Relief lines alongside the terminal it puts me through to Acton Canal Wharf 'box instead of Wembley Mainline. I know it's possible to sidestep this by finding and scrolling through the list of numbers in the display, but it's actually quicker to get down on the ballast and use the old SPT...!

It's the way the BTS is mapped on the system - the BTS will be (should be) mapped to the most appropriate "signal box", because that BTS is providing GSMR coverage for all lines in that location, regardless of how many signal boxes control the lines. We had a similar situation on the East London Line at Highbury & Islington, the BTS is physically located between the NLL and ELL just off the end of Canonbury platform. Because this BTS "is mapped" to Upminster IECC (the controlling signal box for the NLL), any "SG", "Yellow Button" or "Red button" calls from a train on the ELL are routed to Upminster instead of the controlling signal box for the ELL at New Cross Gate. Network Rail offered us a way around this by providing a "1200" number (IIRC "1228") which is a unique short-form number on the GSMR for the New Cross Gate signal box. This number is displayed on a reflective lineside sign (black on white) at H&I. You could try requesting for one of these boards to be provided at Willesden if tge problem persists.

 

Regards, Ian.

[royaloak]

If the GSMR displayed the area code like the CSR used to you would know which box you were connected to and change it if needed, a very good feature they didnt take forward.

[Edwin_m]

Under ERTMS level 2 the system must know which train is where in order to give it a movement authority.  Does the same information get fed into the radio system so it can reliably connect voice calls to the correct signaller? 

[iands]

If the GSMR displayed the area code like the CSR used to you would know which box you were connected to and change it if needed, a very good feature they didnt take forward.

There are no area codes for GSMR, that's a required feature of system - the fact you don't need to keep changing area codes as per CSR is an improvement. The "incorrect routing" is not that a common occurrence (I know someone will now say "it occurrs everytime I stop at "xxx" signal/location), but in the whole scgeme of things "on a daily basis", it is not a regular occurrence. It know it was in certain areas in the early days of GSMR (particularly around the Glasgow area), but improvements to the configuration etc., have since been made - remember this was a completely new system to railways when first introduced and a bit of a steep learning curve for all, but now a lot of those problems have been resolved.

 

Regards, Ian.

[royaloak]

CSR used to automatically update, no need for driver intervention unless you wanted to contact the box covering the area you had just left, a very useful feature.

[iands]

Under ERTMS level 2 the system must know which train is where in order to give it a movement authority.  Does the same information get fed into the radio system so it can reliably connect voice calls to the correct signaller?

 

"Location Dependant Addressing" (LDA) is a feature of GSMR and the location is limited to a "BTS area" (which typically is approx 5km radius). However, an "external" connection to the Train Describer (TD) network allows an "enhanced LDA" feature to present to the signaller (on his/her desktop GSMR terminal) the head code, unit number and signal berth details so that the signaller can see instantly exactly where a call is originating from. So to answer your question, the answer is yes.

 

Regards, Ian.

[iands]

CSR used to automatically update, no need for driver intervention unless you wanted to contact the box covering the area you had just left, a very useful feature.

But CSR didn't always update automatically, hence the reason for all the "CSR Area Code" reminder boards spread about the lineside. GSMR does away with these, a considerable cost saving.

 

Regards, Ian.

[ess1uk]

I’ve heard the Dutch are trying ATO on freight from Rotterdam

Edited January 9, 2018 by ess1uk

[Fat Controller]

I’ve he Dutch are trying ATO on freight from Rotterdam

The 'Betuweroute':-

https://en.wikipedia.org/wiki/Betuweroute

It's been open for some time, but I'm not sure to what extent the ETCS2 has been used. Newer locos, such as the various members of the TRAXX and Vectron, will have been built with at least passive provision for the ETCS equipment, but it would be difficult (and expensive) to retro-fit it to older stock.

[jim.snowdon]

I thought ATO was ETCS level 3?

 

Level 2, ie in-cab signalling, is simpler for freight if you make conservative assumptions about train braking performance and substitute a simple maximum values for train length. The Betuwe line has the advantage, in train control terms, of being a "one traffic" railway, with a narrow range of train speeds. One thing I am not clear about is how ETCS deals with a train becoming divided. Obviously, with automatic air brakes on all vehicles, both parts of the train will come to a stand, but how far apart? Or is this predictable enough to be accounted for by adding a fixed allowance to the maximum train length.

 

Jim

[The Stationmaster]

I thought ATO was ETCS level 3?

 

Level 2, ie in-cab signalling, is simpler for freight if you make conservative assumptions about train braking performance and substitute a simple maximum values for train length. The Betuwe line has the advantage, in train control terms, of being a "one traffic" railway, with a narrow range of train speeds. One thing I am not clear about is how ETCS deals with a train becoming divided. Obviously, with automatic air brakes on all vehicles, both parts of the train will come to a stand, but how far apart? Or is this predictable enough to be accounted for by adding a fixed allowance to the maximum train length.

 

Jim

 

Slight correction Jim

 

'Obviously, with automatic air brakes on all vehicles, both parts of the train should come to a stand ... etc

 

The greatest distance I have known between the two portions - i.e. the rear portion which did stop (only one wagon) and the rest of the train, which obviously kept going, was 12 miles give or take a few chains (and 5 block sections).  although the missing tail lamp was noticed by two of the Signalmen involved we only knew where the missing wagon was because a member of the public 'phoned in and asked if we had lost one because there was one sitting on the railway line at the bottom of his garden.

[Edwin_m]

I thought ATO was ETCS level 3?

 

Level 2, ie in-cab signalling, is simpler for freight if you make conservative assumptions about train braking performance and substitute a simple maximum values for train length. The Betuwe line has the advantage, in train control terms, of being a "one traffic" railway, with a narrow range of train speeds. One thing I am not clear about is how ETCS deals with a train becoming divided. Obviously, with automatic air brakes on all vehicles, both parts of the train will come to a stand, but how far apart? Or is this predictable enough to be accounted for by adding a fixed allowance to the maximum train length.

 

Jim

All levels of ETCS are based on EU specification of functions and key interfaces, so it should be possilble to buy the various bits of equipment needed from different suppliers and they should all work together.  This differs from other train control systems, which generally are produced by a single manufacturer and are not interoperable with other suppliers' equipment. 

 

ATO is Automatic Train Operation, the ability of the train to "drive itself" with the driver just supervising or even not present at all.  It needs an on-board computer to be fed with enough information about the state of the line ahead and its own braking characteristics to select a safe speed.  So in principle any communication system that supplies the train with the necessary information can support ATO.  Various non-interoperable ATO systems are available from different suppliers (as on Crossrail central core, DLR, Northern, Jubilee and Victoria Lines), but ETCS is geared towards the needs of the main line railway rather than the simpler operation found on metros. 

 

ETCS level 3 is a "moving block" system, basically allowing the minimum spacing between trains to depend on their speed and braking capability rather than being fixed by block sections that are defined in hardware.  It also eliminates all trackside equipment other than that associated with points and "dumb" beacons (balises) that just confirm the position to the on-board equipment.  Everything else, including the train advising the system of its own position, is done by radio links.  The main problem with level 3 is that it has been promised for 20 years or so but doesn't actually exist yet (but some of the above proprietary systems do offer moving block). 

 

Nearly every ETCS route has conventional manual driving with the ETCS system providing train protection and sometimes signalling as well.  But ETCS level 2 also provides enough information for ATO and this will be introduced on the central core of Thameslink quite soon.  So ATO is possible without ETCS level 3 and indeed without ETCS at all, and ETCS at any level is possible without ATO. 

[jim.snowdon]

OK, but with freight trains, how does ETCS deal with not knowing where the back end of the train is relative to the front end, since the actual length of a freight train is not predictable? What about the possibility of the train becoming divided, when although the two parts should stop as a result of the loss of brake pipe pressure, the relative stopping positions of the two parts relative to each other are not predictable and the rear sectin, being freight stock, is without any means of reporting its position?

 

I don't doubt that someone has thought of these things, but I am curious to know what the results of that thinking are.

 

Jim

[david.hill64]

OK, but with freight trains, how does ETCS deal with not knowing where the back end of the train is relative to the front end, since the actual length of a freight train is not predictable? What about the possibility of the train becoming divided, when although the two parts should stop as a result of the loss of brake pipe pressure, the relative stopping positions of the two parts relative to each other are not predictable and the rear sectin, being freight stock, is without any means of reporting its position?

 

I don't doubt that someone has thought of these things, but I am curious to know what the results of that thinking are.

 

Jim

Good points.

 

Train integrity detection is an essential requirement for level 3 (and other levels if secondary detection is not provided).

 

Train length information has to be input for any form of ATP as the speed of the train has to be supervised over the whole length of the train, so that for example the rear of the train is not exceeding a restriction when a train accelerates from it.

Edited January 10, 2018 by david.hill64

[peach james]

Wouldn't the FRED provide you with where the back end of the train is?  (at least, if the gap got bigger than when the train started...)  The length of the train is going to be something- Feet, or meters (or KM...), so that data should be easy enough to input along with the destination of the train ?  

 

I'm assuming at some point, someone will have to input data into the computer, with the length, weight, makeup (cars) ect of the train- or that data will be transferred from somewhere else into ETCS for that specific train.

 

From that, the FRED would give the end of train data back to ETCS so that there would be two reporting points for a freight train- the front & the back.  If the data from both ends doesn't reasonably agree, one would expect at ETCS 2 or higher, that it would set signals against and assume the train has split in some form or another.  

 

If you are using ATO, or confuser assisted throttling ( http://getransportation.com/train-performance-improves-ge-transportation%E2%80%99s-digital-solutions-software ) then you would already have speed restrictions in the system, I would expect either to provide the train with guidance to power position required to make the most efficient use of said power.

 

FRED=Flashing Rear End Device, the nasty name for the (usually orange) box on the back of North American trains since 1984 or so...they do provide a bunch more data now than they did in 1983 when they were approved, I understand including velocity & brake status.

 

James

[david.hill64]

Wouldn't the FRED provide you with where the back end of the train is?  (at least, if the gap got bigger than when the train started...)  The length of the train is going to be something- Feet, or meters (or KM...), so that data should be easy enough to input along with the destination of the train ?  

 

I'm assuming at some point, someone will have to input data into the computer, with the length, weight, makeup (cars) ect of the train- or that data will be transferred from somewhere else into ETCS for that specific train.

 

From that, the FRED would give the end of train data back to ETCS so that there would be two reporting points for a freight train- the front & the back.  If the data from both ends doesn't reasonably agree, one would expect at ETCS 2 or higher, that it would set signals against and assume the train has split in some form or another.  

 

If you are using ATO, or confuser assisted throttling ( http://getransportation.com/train-performance-improves-ge-transportation%E2%80%99s-digital-solutions-software ) then you would already have speed restrictions in the system, I would expect either to provide the train with guidance to power position required to make the most efficient use of said power.

 

FRED=Flashing Rear End Device, the nasty name for the (usually orange) box on the back of North American trains since 1984 or so...they do provide a bunch more data now than they did in 1983 when they were approved, I understand including velocity & brake status.

 

James

In Europe there is no requirement for a FRED: just a continuous brake and a tail lamp. BR did try something similar in conjunction with automatic brake continuity testing of coal trains.

 

The point about ETCS is that unlike some other systems (eg Shinkanshen and many metros) there is no need for the onboard signalling to have a map of gradients, speed restrictions etc. The balise telegram provides the onboard system with the information it needs to know regarding safe speed, restrictions and gradients. The onboard could use this discrete data to optimise driving and that maybe what is planned for Thameslink, but I don't have details of the implementation.

[Steadfast]

OK, but with freight trains, how does ETCS deal with not knowing where the back end of the train is relative to the front end, since the actual length of a freight train is not predictable?

Jim

Presumably similar to now, when the train departs it's origin, the train length is entered into the trains computer (Qtron) by the driver using the information off the trainlist paperwork

 

Jo

[D1059]

These usually occur because something has changed since the original radio survey and build was undertaken - like a new building(s) sprouting up, lineside vegitation growing unchecked for several years, 

 

If lineside vegetation can cause a network blackspot I'm surprised you can get a signal anywhere

[iands]

If lineside vegetation can cause a network blackspot I'm surprised you can get a signal anywhere

You would be surprised just how much lineside vegetation could affect radio coverage, especially after a heavy/prolonged downpour. It depends where the train is in relation to the base station at the time of 'radio transmission', but the effects could be quite significant - and then within about 10-15 minutes, it could all change again and you wouldn't know there had been a problem in the first place! Radio can be a very peculiar thing at times.

 

Regards, Ian.

[Christopher125]

It is virtually impossible to fit to various preserved locos, especially those propelled by steam as it not only initiates a brake application in certain conditions if the Driver doesn't do it but will so cut off the power.  Not to difficult to do things with brake systems (apart from all sorts of graduated applications etc) - after all the GWR was doing it over a century back.  But cutting off the power is a rather more sophisticated task which means going into the control system of the loco and it would not be at all easy to incorporate on a steam engine without some design changes of a very basic sort.

 

Incidentally there is another challenge with freight when it comes to braking as train load has to be taken into account plus the way in which brake force develops through the train.  BR trials with ATP showed it was well nigh impossible to get consistent brake application on a freight, even using the same train several times over which came up with a different stopping distance every time.  Freight fitment could prove to face some interesting hurdles I think.

 

IIRC the A1 Trust's Rob Morland is confident that there are no fundamental obstacles for steam if they have sufficient power and room, their Autumn 2015 newsletter "confirmed that Interfleet Technology had approved our scheme for fitting ERTMS signalling equipment with an application date now fixed for 2019"

 

As for freight, ETCS is already used by freight trains on various lines in Europe and beyond - the Betuweroute and new Gotthard Tunnel being obvious examples.

 

OK, but with freight trains, how does ETCS deal with not knowing where the back end of the train is relative to the front end, since the actual length of a freight train is not predictable? What about the possibility of the train becoming divided, when although the two parts should stop as a result of the loss of brake pipe pressure, the relative stopping positions of the two parts relative to each other are not predictable and the rear sectin, being freight stock, is without any means of reporting its position?

 

What's proposed (and already in wide use) is ETCS Level 2 which still involves blocks with axle counters or track circuits, but without the constraints of lineside signals.

 

While Level 3 has yet to be developed, it would also do away with track-based train detection - this presents obvious difficulties for freight trains and confirming their integrity.

 

However there is a hybrid solution under consideration, as described in the article below - this would retain blocks with track-circuits or axle-counters, but with ‘virtual blocks’ within them. 

 

The Rail Engineer: ERTMS Level 3 - A possible way forward

Edited January 11, 2018 by Christopher125

[david.hill64]

Bombardier has a system in Kazakhstan which is essentially ETCS level 3, except that it uses TETRA radio as the transmission medium. Special End of Train detection units were also supplied by Bombardier for the application which I think is the only application of moving block technology to a mainline railway.