• Our booking engine at tickets.railforums.co.uk (powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!

Cost of electrification OHLE vs 3rd rail

Status
Not open for further replies.

DarloRich

Veteran Member
Joined
12 Oct 2010
Messages
29,182
Location
Fenny Stratford
Only trying to pose the case for 3rd rail, doesn't seem to make much sense as to why it is in such wide use if it is completely unsuitable. Perhaps you can link some data on running costs for 3rd rail vs diesel traction before you sigh?

I don't have to offer any information on running costs. You ask about the case for third rail: there isn't one. It is an old fashioned,outdated approach to providing electrification for a modern, 21st century railway. There is your answer.

PS go and read accident reports about staff electrocuted by 3rd rail and come back and tell me how great it is.
 
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R

RailUK Forums

hwl

Established Member
Joined
5 Feb 2012
Messages
7,354
Only trying to pose the case for 3rd rail, doesn't seem to make much sense as to why it is in such wide use if it is completely unsuitable. Perhaps you can link some data on running costs for 3rd rail vs diesel traction before you sigh?

History (ignoring LU) - the DC electrical equipment could be supplied by multiple British companies whereas the original AC OHLE equipment for LBSCR (1909 onwards South London electrification*) inc on train traction gear was supplied by Siemens (Munich) or AEG (Berlin) and there were some unforeseen equipment and spares supply issues from July 1914 onwards that eventually resulted in the LBSCR OHLE being converted to match the neighbouring LSWR and SER 3rd rail post 1923 grouping.

*The overhead wires at Balham Jn (looking north) can be seen being erected in this photo:
https://esngblog.files.wordpress.com/2016/12/image1-online11.jpg

Where it is already there it makes sense to continue including extensions if they can be carefully managed but no large new schemes)

AC OHLE only then started to be contemplated in the UK again in the mid - late 1950s when France and the UIC was heading towards 25KV 50Hz as the future so AC wasn't just a single factory German monopoly, but multiple western suppliers e.g. AEI, Alstom, Asea, Brown - Boveri and EE too (the AEG factory ending up in East Berlin, now Bombardier Transportation's HQ).

BR's haste to get rid of steam by an arbitrary date meant that 3rd rail was the only deliverable solution for Bourenmouth by the steam withdrawal deadline when the choice was made in 1965. If BR had held on to steam for another 18months the outer SWML would have been 25kV.
 
Last edited:

AM9

Veteran Member
Joined
13 May 2014
Messages
14,191
Location
St Albans
This is however not including any of the possible mitigation measures that can be implemented.
A substantial fraction of these are things like electric shock from people falling off station platforms - which can be mitigated producing pseudo-gaps in station platforms where the centre portion of the platform rail is only livened if necessary because a train can't get a shoe on the live component.

That would be a major investment at every 3rd rail station on the network. It is unproven in practice, and attaining a reliability adequate for even a 3rd rail station would potentially be subject to years of development in use. It could only be justified during a project to replace life-expired 3rd rail infrastructure and would likely cost more to implement that an upgrade to 25kV OLE.

Also improved signage could be argued to reduce the tresspasser fatalities which are a major portion of the risk generated by third rail instalaltions. As could providing insulation on the sides of the conductor rail so only the top surface is actually a contact hazard.

The you have been warned defence is no longer acceptable, especially where the cost of a proven safer method of electrification is within the limits of 'reasonable'.

Typical DC losses including very high density areas where volt drop in the rails will be huge and contribute to losses.
These are not regions typical of projected expansions to the electrification system.

The losses are great in metro areas because of the very heavy currents involved in running high density services. In rural areas, the the legnth of lines becaomes the problem. Feeding a 3MW load at the end of miles of steel rails still requires track feed points quite close to each other. All of the capital equipment and it's high cost of ownership along a moderately busy mainline would make electrification (by 3rd rail) a very bad business case. To electrify a line like Manchester to Leeds with 3rd rail would be insane: heavy trains, significant gradients, ice issues, and very remotely located unprotected ground-level conductors would be rejected outright on first application.

Also why are AC systems allowed to return power to the grid from regeneration but DC systems are not?
Modern AC-DC converter stations are easily capable of this.
Have to compare apples toa pples.

Because a) as hwl says, with 3rd rail trains, little if any regenerated energy gets as far as the substation because it is either used by other trains or merely disappears into the series resistance of the track and feeders
and b) with ac, there is no need for complex converters at feed points, - any passive transformer is a bi-directional device and will accept the regen energy for returning to grid. The electronics to generate an in-phase return ac waveform (or an increased DCV) is inherent in the design of the traction control equipment fitted to all current EMU designs.
 

Meerkat

Established Member
Joined
14 Jul 2018
Messages
7,420
If the LBSCR was overhead does that mean their is headroom to make it OHLE again?
 

hwl

Established Member
Joined
5 Feb 2012
Messages
7,354
If the LBSCR was overhead does that mean their is headroom to make it OHLE again?
No it was 6.6kV and they didn't really believe in clearance in the modern sense for bridges* e.g. West Norwood station over bridge where you can still the hammer + chisel work and where the contact wire cut into the brick. (I can't find a photo on line immediately)

* "Brick is non-conductive in'it"
 

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
16,635
That would be a major investment at every 3rd rail station on the network. It is unproven in practice, and attaining a reliability adequate for even a 3rd rail station would potentially be subject to years of development in use. It could only be justified during a project to replace life-expired 3rd rail infrastructure and would likely cost more to implement that an upgrade to 25kV OLE.
Except the costs of 25kV have been demonstrated by actual projects is astronomical.
A normally dead section of rail in the centre of a station platform doesn't require any new technologies to be developed - nor does it have to be safety grade protection because at no time would the rail be assumed to be dead.
It working reliably 90% of the time still reduces the risk of people falling onto the platform and getting a shock by 90%......

The you have been warned defence is no longer acceptable,
It's not the "you have been warned defence"
It's the "people are less likely to tresspass due to better signage, so the risks are lower" defence.
If better signage reduces FWI occurences by 20%, the value of other engineering works to reduce those injuries falls by about 20%.

especially where the cost of a proven safer method of electrification is within the limits of 'reasonable'.
If by reasonable you mean "so high that the electrification programme has been killed by them".

The losses are great in metro areas because of the very heavy currents involved in running high density services. In rural areas, the the legnth of lines becaomes the problem. Feeding a 3MW load at the end of miles of steel rails still requires track feed points quite close to each other.
Why are we feeding through steel rails?
The conductor rails would be made of aluminium with a thin steel wear surface, as is used in most third rail systems in the world, and most of the return current would be flowing through aluminium cables connected in parallel with the running rails?

High cross section low voltage cables made of aluminium are really cheap compared to the cost of anything else on the railway.

This is not the 1920s or the 1950s.

Also in what rural environment would the traction load be 3MW?
Even a Class 444 is only 2MW.....

A more likely load is about 1-1.5MW
Even two 3-car Class 377s is going to be way less than 2MW.

EDIT:
Even a 5-car Class 802 only has 2.1MW of engine output.....

All of the capital equipment and it's high cost of ownership along a moderately busy mainline would make electrification (by 3rd rail) a very bad business case. To electrify a line like Manchester to Leeds with 3rd rail would be insane: heavy trains, significant gradients, ice issues, and very remotely located unprotected ground-level conductors would be rejected outright on first application.
Who said anything about Manchester to Leeds?
Given the capital equipment costs of a railway project of the type (more like Marshlink or WEML) I am recommending are believed even by Network RAil to be drastically lower than the 25kV solution... thats a hell of a lot of cost of ownership!

Because a) as hwl says, with 3rd rail trains, little if any regenerated energy gets as far as the substation because it is either used by other trains or merely disappears into the series resistance of the track and feeders
and b) with ac, there is no need for complex converters at feed points, - any passive transformer is a bi-directional device and will accept the regen energy for returning to grid. The electronics to generate an in-phase return ac waveform (or an increased DCV) is inherent in the design of the traction control equipment fitted to all current EMU designs.

You mean "complex converters" like those currently being installed as part of the ECML PSU2 upgrade?
AC Regeneration is not nice for the grid at all, since a train is inherently a single phase load and the grid really doesn't like single phase loads.

Power electronics in the railway power distribution system are the future.
 
Last edited:

najaB

Veteran Member
Joined
28 Aug 2011
Messages
30,692
Location
Scotland
Except the costs of 25kV have been demonstrated by actual projects is astronomical.
The massive overruns on the GWML project were mostly caused by poor planning and design work rather than due to the cost of the civil engineering. Anything will be twice as expensive as possible when you have to do the work twice!
 

apk55

Member
Joined
7 Jul 2011
Messages
438
Location
Altrincham
750V DC systems are really only suitable for train loads of about 2MW although they have in places stretched this to 4MW at great expense with substations every mile. 25KV AC systems can easily deal with loads of 8MW or more and this is with substation distances of 10 to 15miles. This makes it unsuitable for fast intercity passenger trains or freight trains. And many suburban services share tracks with these services at least for part of the journey so a standard system is required that can meet all needs.
The unbalanced single phase loading of the grid can be addressed by the introduction of inverter substations which can convert a single phase poor power factor load into a balanced 3 phase load.
 

hwl

Established Member
Joined
5 Feb 2012
Messages
7,354
The unbalanced single phase loading of the grid can be addressed by the introduction of inverter substations which can convert a single phase poor power factor load into a balanced 3 phase load.
Already being done in slightly different ways with Autotransformers for AC OHLE.

3rd rail supplies have been 3 phase derived for over 100 years.
 

AM9

Veteran Member
Joined
13 May 2014
Messages
14,191
Location
St Albans
The massive overruns on the GWML project were mostly caused by poor planning and design work rather than due to the cost of the civil engineering. Anything will be twice as expensive as possible when you have to do the work twice!
I know, the 'GW or NW electrification schemes ran massively over budget' chorus has been touted so much on here, it's beginning to sound like some of the misconceptions trotted out on another far bigger political issue in the UK. Then comes the fallacious conflation of an argument that because those two projects overspent, then all OLE schemes will similarly overspend.
As you say, the GWR electrification went awry in the first instance because it was a rebooting of the expertise to plan, procure and oversee what turned out to be a massive (re)learning curve. Then there were the issues like the S&T cables being hurriedly covered years before to prevent cable theft with no records kept. Consequently, when the boring for piles started there were several serious interruptions to maionline services because the drills had damaged the hidden cables, so all boring had to be done manually. That also prevented the benefits of the High Output Plant System (HOPS) train from being used. None of this is directly related to installing OLE over a line, - more issues about the UK's start stop relationship with electrification followed by an appalling lack of proper records.
The failure of inexperience also appeared again on the Manchester to Bolton section of the NW scheme, where insufficient knowledge of the underlying mining heritage meant successive setbacks on OLE infrastructure.
Putting OLE up across open countryside can be relatively inexpensive, e.g. the mentioned Marshlink is about 27 miles long. There are to my count about 10 bridges over the track of which four are quite modern so may not require expensive (or any) modifications to accommodate OLE, one tunnel - less than 3/4 of a mile long and a couple of footbridges. All other road crossings are level so no problems there. An end feed of 25kV for , the line from Ashford would probably present no problems whatever EMUs were run on the line but there would need to be quite a lot of DC feed infrastructure to make through services to London viable. As a through-life cost (say 50 years, ac would probably have far better business case, and of course with all those level crossings, a lower risk of electrocution that high voltage lines over three times as high as any adults have.
 
Last edited:

hwl

Established Member
Joined
5 Feb 2012
Messages
7,354
Why are we feeding through steel rails?
The conductor rails would be made of aluminium with a thin steel wear surface, as is used in most third rail systems in the world, and most of the return current would be flowing through aluminium cables connected in parallel with the running rails?

High cross section low voltage cables made of aluminium are really cheap compared to the cost of anything else on the railway.

The problem with aluminium is the rail length is limited to extrusion press x die geometry length which happens to be 18m at which point you have to fishplate join the rail together resulting in losses.
It also has to be stainless clad for corrosion reason and the conductivity of stainless is worse than normal steel.

NR found a very good alternative going heavy section on 3rd rail and running rails for return current. And going for flash butt welding as the preferred 3rd rail joining method to reduce losses.
LU have gone for the aluminium conductor rail in places as they have more geometry issues (especially 4th rail centre rail) than make the new extra heavy section steel conductor rail less practical for them.
 

hwl

Established Member
Joined
5 Feb 2012
Messages
7,354
A normally dead section of rail in the centre of a station platform doesn't require any new technologies to be developed - nor does it have to be safety grade protection because at no time would the rail be assumed to be dead.
It working reliably 90% of the time still reduces the risk of people falling onto the platform and getting a shock by 90%......

What if people learnt they are switched off outside of train running times in some places and start going for a wonder across the tracks as it is quicker/ avoids the RPI on the footbridge etc. in others and get zapped?
 

Wychwood93

Member
Joined
25 Jan 2018
Messages
634
Location
Burton. Dorset.
What if people learnt they are switched off outside of train running times in some places and start going for a wonder across the tracks as it is quicker/ avoids the RPI on the footbridge etc. in others and get zapped?
The juice rail, even on Christmas Day, remains live - unless there is a reason for it not to be (engineering stuff etc.).
 

hwl

Established Member
Joined
5 Feb 2012
Messages
7,354
The juice rail, even on Christmas Day, remains live - unless there is a reason for it not to be (engineering stuff etc.).
Yep - I was just pointing out that various proposals from those outwith the 3rd rail area to improve safety by turning it off may not have the desired result!
 

Meerkat

Established Member
Joined
14 Jul 2018
Messages
7,420
As soon as people realised parts of the third rail weren’t electrified those parts would go missing in the night!
 

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
16,635
What if people learnt they are switched off outside of train running times in some places and start going for a wonder across the tracks as it is quicker/ avoids the RPI on the footbridge etc. in others and get zapped?

How would they even get that information unless they go around reading technical reports from the RSSB?
That sort of information is not something the types of people who indulge in such behaviours care about.

The problem with aluminium is the rail length is limited to extrusion press x die geometry length which happens to be 18m at which point you have to fishplate join the rail together resulting in losses.
It also has to be stainless clad for corrosion reason and the conductivity of stainless is worse than normal steel.
The conductivity of the stainless steel is essentially irrelevant, since there is only a few millimetres of in the conductor path between the shoe and the bulk aluminium portion of the rail.

Effetively no current flows along the rail in the stainless steel portion.

Fishplate losses can be mitigated using aluminium fishplates and similar components that drastically reduce the loop resistance.


NR found a very good alternative going heavy section on 3rd rail and running rails for return current. And going for flash butt welding as the preferred 3rd rail joining method to reduce losses.
LU have gone for the aluminium conductor rail in places as they have more geometry issues (especially 4th rail centre rail) than make the new extra heavy section steel conductor rail less practical for them.

It's going to have to be a very heavy section to get anywhere near the conductivity of aluminium.....

And then theres the fact that if you can make the steel rail heavier.... you can make the aluminium rail heavier too.
EDIT:

Perhaps Network Rail truly is correct, and they are a waste of time, but the fact that virtually everyone else, including the London Underground, are going for composite conductor rails is more than slightly peculiar.
 
Last edited:

hwl

Established Member
Joined
5 Feb 2012
Messages
7,354
Fishplate losses can be mitigated using aluminium fishplates and similar components that drastically reduce the loop resistance.




It's going to have to be a very heavy section to get anywhere near the conductivity of aluminium.....

And then theres the fact that if you can make the steel rail heavier.... you can make the aluminium rail heavier too.
EDIT:

Perhaps Network Rail truly is correct, and they are a waste of time, but the fact that virtually everyone else, including the London Underground, are going for composite conductor rails is more than slightly peculiar.

Aluminium forms a natural (insulating) oxide layer*, the local environment influences how thick and the precise chemistry so the joins are more lossy than you might think no matter how big the fishplates. (*chose the right acids + some current and it is called anodising!)
Hence key to try to keep moisture out of the joins which is easier indoors in a tunnel.

Section wise upto 75kg/m.

The big difference between LU and NR is that on NR the return current goes through the running rails (x2 equivalent to 112.6kg/m section) but on LU via the single 4th rail hence conductor rail impedance being more important on LU than NR when the return path is factored in too.
 
Last edited:

Bald Rick

Veteran Member
Joined
28 Sep 2010
Messages
29,070
The massive overruns on the GWML project were mostly caused by poor planning and design work rather than due to the cost of the civil engineering. Anything will be twice as expensive as possible when you have to do the work twice!

The initial estimating wasn’t great either.

No one ever mentions the cost overruns on Edinburgh - Glasgow, which was by some margin the most expensive recent electrification project in cost per km.
 

Meerkat

Established Member
Joined
14 Jul 2018
Messages
7,420
The initial estimating wasn’t great either.

No one ever mentions the cost overruns on Edinburgh - Glasgow, which was by some margin the most expensive recent electrification project in cost per km.

Can’t be right, Scotland is apparently how all railways should be run :D
 

mcmad

Member
Joined
11 Mar 2015
Messages
977
Is that the pure electrification cost or including things like slab tracking Winchburgh and Queen St tunnels?
 

Bald Rick

Veteran Member
Joined
28 Sep 2010
Messages
29,070
Is that the pure electrification cost or including things like slab tracking Winchburgh and Queen St tunnels?

The whole job, but then GW electrification had lots of difficult parts too (clearing Box tunnel for example). Indeed almost every electrification project has difficult components. Edinburgh - Glasgow was by no means out of step in that respect. Indeed it was amongst the cheapest in terms of route clearance (where there was any).

But as it happens, Edinburgh - Glasgow was nearly the most expensive for the OLE works as well.
 
Last edited:

Tio Terry

Member
Joined
2 May 2014
Messages
1,178
Location
Spain
One of the problems with bi-metallic conductor rail is differences in rate of expansion of the two metals which, in extreme cases, can lead to them parting. It’s much better suited to tunnels (LU) where temperature variations are much more limited.
 

LNW-GW Joint

Veteran Member
Joined
22 Feb 2011
Messages
19,555
Location
Mold, Clwyd
What was the last 3rd rail project in the UK? I’m thinking Hastings or East Grinstead electrification. It seems very short sighted that no others will be done. Uckfield and North Downs line are obvious candidates

The Merseyrail 3rd rail system was extended from Hooton to Chester and Ellesmere Port in 1993/4.
Other extensions have been considered since but none approved, and they will probably be built with AC equipment.
 

apk55

Member
Joined
7 Jul 2011
Messages
438
Location
Altrincham
Practically all new 3rd rail stock is designed with conversion to 25KV in mind with places to put a pantograph, transformer and rectifier (well 4 quadrant converter) allocated. Only exception is probably LU stock.
So extensions of the network on 25kv should not be a problem.
One of the big problems on 3rd rail systems is return voltage drop. More than a few volts and you run into safety and corrosion issues. Touch a running rail or something connected to it such as a train and a remote earth and you could get a shock. This is one advantage of LU 4 rail system where the return path is isolated from the environment.
Running rails are made of wear resistant steel which is considerably higher resistance than the high conductivity steel used for the 3 rail so return voltage drop is a big problem.
 

hwl

Established Member
Joined
5 Feb 2012
Messages
7,354
Practically all new 3rd rail stock is designed with conversion to 25KV in mind with places to put a pantograph, transformer and rectifier (well 4 quadrant converter) allocated. Only exception is probably LU stock.
So extensions of the network on 25kv should not be a problem.
One of the big problems on 3rd rail systems is return voltage drop. More than a few volts and you run into safety and corrosion issues. Touch a running rail or something connected to it such as a train and a remote earth and you could get a shock. This is one advantage of LU 4 rail system where the return path is isolated from the environment.
Running rails are made of wear resistant steel which is considerably higher resistance than the high conductivity steel used for the 3 rail so return voltage drop is a big problem.
The combined running rail resistance is only 20% more than the 3rd rail resistance especially with NR now using heavier section running rail to help in 3rd rail areas.
The worst case voltage drop means you can get nasty shock of the running rails (more US domestic mains rather the European mains voltage type shock) albeit it is quite hard to achieve those conditions.

It is worth bearing in mind the impedance of running rails for AC return current is very much worse than the DC resistance /impedance with the magnetic permeability of the running rails increased due to the useful alloying elements in the running rail in the AC case in addition to skin effect issues.
 

LNW-GW Joint

Veteran Member
Joined
22 Feb 2011
Messages
19,555
Location
Mold, Clwyd
The initial estimating wasn’t great either.
No one ever mentions the cost overruns on Edinburgh - Glasgow, which was by some margin the most expensive recent electrification project in cost per km.

The list I saw had the Manchester-Blackpool costs as higher even than the E&G (in £/stkm).
But it's not clear if that included the Blackpool remodelling/resignalling scheme or the troublesome Farnworth re-tunnelling.
The project essentially installed two lots of foundations over much of the distance, with multiple design changes along the way.
By contrast, the initial Manchester-Edge Hill wiring (including the Chat Moss section), which used "traditional methods", came in fairly close to plan (time and cost).
 

a_c_skinner

Established Member
Joined
21 Jun 2013
Messages
1,576
For me the issue isn't that we should have the ability to have new 3rd rail schemes but that some pretty tempting schemes (Marshlink for example) within existing areas of 3rd rail are apparently ruled out totally. Isn't Uckfield deeply costly for OHLE because of a restrictive tunnel? Forbidding that seems needlessly restrictive.
 

najaB

Veteran Member
Joined
28 Aug 2011
Messages
30,692
Location
Scotland
For me the issue isn't that we should have the ability to have new 3rd rail schemes but that some pretty tempting schemes (Marshlink for example) within existing areas of 3rd rail are apparently ruled out totally.
New 3rd rail isn't ruled out completely. Anyone proposing a scheme just has to provide a justification strong enough to override the presumption against approval.
 
Status
Not open for further replies.

Top