Grade Separation?

[Trog]

However, the accident report into the August 1996 Watford Junction accident makes for an interesting read (a head on collision on a flat crossover resulting from a SPAD, with the specific design of the signalling and speed profile being one of the contributory factors).

Grade separating Watford South, Watford Station and Watford North Junctions would have been quite a challenge. Even now with the Watford Station ladder gone it would be very difficult.

 

[30907]

The Weybridge one is actually know as the Byfleet Curve (Diveunder on the Down from Addlestone that joins the Down Slow Line immediately before West Byfleet station. This is an old diagram where the station was still known as West Weybridge. I wonder why at Weybridge the junction was never grade separated coming off the Down Slow heading towards Addlestone?

Lack of space plus lack of need: a handful of rush hour trains in the evening peak were all that used the connection in both 1910 and 1938 - whereas the Byfleet Curve was a vital freight link from Feltham and the North. The Frimley curves were likewise never grade separated, even though the East curve was electrified and carried passenger services.

 

[MidlandMainlie]

An obvious place on the rail network that should be grade separated is the flat crossing at Newark, perhaps we might find out when the East Coast RUS is published?

 

[QueensCurve]

Didcot comes up as a key future flyover location.
Ordsall Chord should be, but there is no space - so flat junctions.
Bletchley was one of the abortive ones (which should be coming back with E-W Rail).
The freight lines through Crewe ("Independent Lines") are a long-standing example.

Nobody seems to have meantioned Weaver Junction.

 

[Topgun333]

Grade separation needs a plain English term. Highways England use the term to describe proposed new junctions at public exhibitions and it confuses many people. I think elevated junction or split level junction is easier to understand.

 

[Trog]

Elevated junction is that where you build a roundabout on top of a little hill, to make it easier for people to slow down approaching it and accelerate away?

 

[Senex]

Grade separation needs a plain English term. Highways England use the term to describe proposed new junctions at public exhibitions and it confuses many people. I think elevated junction or split level junction is easier to understand.

What's wrong with the long-established flying junction or burrowing junction? Or perhaps non-conflicting junction? Grade-separated junction seems to be an import from road construction and American English.

 

[najaB]

What's wrong with the long-established flying junction or burrowing junction? Or perhaps non-conflicting junction? Grade-separated junction seems to be an import from road construction and American English.

'Grade separated' covers both flying and burrowing with the one term.

 

[MarkyT]

'Grade separated' covers both flying and burrowing with the one term.

Agreed it is a generic modern term widely used in civil engineering.
In many cases when design begins the client doesn't don't know or care much which line or road carriageway will burrow under or fly over which, only that they need to be separated to handle the required traffic volume without conflict. Reading remodelling was a good recent example of this. The main line dive-under envisaged originally morphed very quickly into a flyover alternative in detailed design once it became clear that would be much easier to build.

 

[Hartington]

Going back to the reason for the original post the OP seemed to suggest either 4 tracking or grade separation. That conjured up images of a double deck 4 track railway!

Would grade separating junctions really produce the same capacity improvement as 4 tracking?

 

[najaB]

Would grade separating junctions really produce the same capacity improvement as 4 tracking?

Naturally, the answer is: depends.

If the majority of traffic is along one route, but it's a mixture of slow and fast services (stopping/express or passenger/freight) then quad-tracking will probably be the solution as it allows faster services to overtake slower ones. If, on the other hand, there is a diverging route that is heavily used (either used frequently or used by slow/long trains) that will benefit from grade separation.

 

[route:oxford]

Going back to the reason for the original post the OP seemed to suggest either 4 tracking or grade separation. That conjured up images of a double deck 4 track railway!

Would grade separating junctions really produce the same capacity improvement as 4 tracking?

There are already 4 tracks at Diddyland.

Grade separation simply allows you to get from one line to another without using the line(s) inbetween.

So going to Didcot as an example. A service travelling from London to Oxford on the "fast down".

On Google Maps, you'll see that there is a bridge to the East of Tesco. At this point the London train has to move from the "fast down" onto the "fast up" and travel along it until it reaches Tesco.

It then moves from the "fast down" to the "relief up"

It then continues on the "relief up" until it reaches Aldi, then it moves over to the Didcot avoiding "down" line.

So for the time that it undertakes that move, the only other movement that can happen on the signalling gap between Didcot Station and the bridge East of Tesco is a service on the "Up Avoiding" line to the "relief up" line.

 

[Ships]

Isn't Aynho junction already grade separated?

 

[Senex]

Isn't Aynho junction already grade separated?

Always has had a flyover for the down Bicester.

 

[hwl]

Nobody seems to have mentioned Weaver Junction which was the first.

Only if you believe Wikipedia...

I thought Norwood Fork Jn was the first grade separated rail crossing the world opening in November 1845? (The pneumatic railway had it tracks outside the LBSCR/SER shared lines that are now the East Croydon - London Bridge fast lines so had to have a grade seperated crossing at Norwood Fork to get to West Croydon. This helped create the current paired by direction layout on the route. The replacement structure is still in use carrying the slow lines from Norwood Jn to West Croydon.)

And the LBSCR also had 2 others in the area which pre-dated Weaver by ~15 years. Sydenham Jn 1854 (London Bridge to Crystal Palace opening on the same day as the re-sited Crystal Palace) and Norwood North/Bromley Jn flyover (Crystal Palace - Norwood Junction 1856 over which the Victoria - Brighton trains ran from 1856-1866 before the newer route via Selhurst opened).

 

[swt_passenger]

Isn't Aynho junction already grade separated?

I think there was some confusion around posts 12 & 13. The question as I understood it was "which key junctions need grade separation" meaning additional key junctions, however it was answered as a short list of existing key junctions...

 

[QueensCurve]

Only if you believe Wikipedia...

I thought Norwood Fork Jn was the first grade separated rail crossing the world opening in November 1845?

I don't, and stand corrected.

 

[snowball]

Following the success of Norton Bridge, Acton and London Bridge - Network Rail are consulting on Werrington, just north of Peterborough linking the GNML and the GN&GE.

https://www.networkrail.co.uk/east-coast-main-line/werrington-junction/

The consultation material that was supposed to appear on the web on the 7th now finally has. Click on the link above and go near the bottom of the page, just above the heading "Online feedback".

 

[Grumpy]

The consultation material that was supposed to appear on the web on the 7th now finally has. Click on the link above and go near the bottom of the page, just above the heading "Online feedback".

Looking on Real Time trains for yesterday from 7am till close of play there would have been 16 trains, of which 6 freight, that could have used the dive-under. Whilst not disputing this would be a "nice to have" surely there must be busier and more deserving locations to spend the money

 

[The Planner]

You are looking at it now, not 20 or 30 years into the future.

 

[MarkyT]

You are looking at it now, not 20 or 30 years into the future.

Exactly, the current flat junction imposes a limit on the number of conflicting movements. Simply looking at numbers of existing movements tells nothing about suppressed demand, nor performance implications.

 

[Grumpy]

Exactly, the current flat junction imposes a limit on the number of conflicting movements. Simply looking at numbers of existing movements tells nothing about suppressed demand, nor performance implications.

I accept these comments, however they might also apply to other potential locations.
I should also confess that my earlier post only quoted the number of down trains but as I now realise a two track dive-under is proposed the total may increase.

 

[edwin_m]

The consultation leaflet shows that the diveunder will pass beneath the Up Slow, Up Fast and Down Fast of the ECML to link with the Stamford lines to the west (the ECML Down Slow is combined with the Down Stamford as the westernmost track). All freight to and from Felixstowe uses these Stamford lines through Peterborough (because they continue under the ECML towards Ely) so any using the Joint line to or from the north currently has to cross the ECML on the flat somewhere between Peterborough and Werrington.

This means that a gap of a few minutes is needed on each of those three ECML tracks whenever a freight goes to or from the Spalding route, and therefore that re-routeing it via the diveunder creates the potential for three more trains on the ECML. The benefit is even more if the diveunder allows more freight trains to use the Joint line instead of the ECML.

There is a further small benefit if the passenger service to Spalding and beyond makes use of the diveunder and terminates in one of the western platforms at Peterborough. This could even facilitate London-Spalding trains with any future electrification - something that has been talked about for decades though probably still a long way off. The Down trains would use a western platform at Peterborough and the diveunder, and the return workings would join the Up Slow to use an east side platform.

 

[MarkyT]

I accept these comments, however they might also apply to other potential locations.

That's a fair comment. However a significant reason why this work represents particularly good value is that it allows most freights to be removed from the ECML double track sections entirely south of Doncaster, where their relatively low speed and poor acceleration compared to other traffic severely constrains service frequency for all classes of trains and scheduling flexibility. Allowing routine flexible freight running at almost any time via the 'Joint' line among slower low frequency regional passenger trains, is almost equivalent to adding a second pair of tracks to the route throughout, with the grade separation providing access involving zero conflict with the expresses. This project is primarily about speed segregation on a major trunk corridor, not merely solving a few junction conflicts. Now you might say BR shouldn't have closed March- Spalding, and you might be right, but we are where we are and it's not practical to restore that route today. Besides, grade separation at Werrington Jn represents a very near equivalent recreation of that route for freight and also offers some useful relief of conflict for passenger trains around the Peterborough area. It is, in short, a very valuable scheme and it's no surprise NR wishes to proceed with it as a matter of urgency.

 

[edwin_m]

The various junctions around Doncaster also provide some good grade separation for the northern end of the Joint line.

 

[rebmcr]

When grade-separating quad tracks paired by direction, is it ever considered to slew the outer tracks and insert the turnouts in-between? Example an alternate Hitchin could be: DS-turnout-DF // UF-turnout-US, instead of turnout-DS-DF // UF-US-turnout.

That would remove the need for a DF train to cross the DS in order to access the diverging route, and require not much more land take than for an 'outer' turnout. Also, since the slewed outer tracks are likely to be the Slows, it probably wouldn't cause much compromise on linespeeds.

 

[snowball]

When grade-separating quad tracks paired by direction, is it ever considered to slew the outer tracks and insert the turnouts in-between? Example an alternate Hitchin could be: DS-turnout-DF // UF-turnout-US, instead of turnout-DS-DF // UF-US-turnout.

That would remove the need for a DF train to cross the DS in order to access the diverging route, and require not much more land take than for an 'outer' turnout. Also, since the slewed outer tracks are likely to be the Slows, it probably wouldn't cause much compromise on linespeeds.

Quite a lot of grade-separated junctions have one or two tracks round the outside in this way.

 

[Senex]

When grade-separating quad tracks paired by direction, is it ever considered to slew the outer tracks and insert the turnouts in-between? Example an alternate Hitchin could be: DS-turnout-DF // UF-turnout-US, instead of turnout-DS-DF // UF-US-turnout.

Wasn't an earlier plan for Hitchin exatcly that on the down side, with 70-mph turnouts from both DS and DF into the flyover line between them?

 

[MarkyT]

Wasn't an earlier plan for Hitchin exatcly that on the down side, with 70-mph turnouts from both DS and DF into the flyover line between them?

There are very few non-stop trains on the slows through Hitchin, so in real time the down stoppers, whether going to Huntingdon or Letchworth can be fairly easily regulated in the platform to allow a Cambridge fast to overtake or cross the Down Slow in front of them and gain access to the flyover. A small slither of pathing time can be added into stopping schedules precisely to assist this kind of thing. Diverting the down slow around the outside of the flyover line would have been possible I'm sure, but costly and presumably deemed not worth the expense in this particular case.

 

[Chris125]

The double grade separation at Hampton Court Junction is impressive and in fact it can be easy to forget about the line to/from Oxshott given the flyover taking the Down Hampton Court branch. This is an old diagram that appears to show the Down Hampton Court branch as being a dive-under.

I can't see any evidence of a dive under on the Hampton Court branch - you can see on the NLS mapping site before and after the flyover was built.