Should the third rail ban be lifted?

[Falcon1200]

For people on foot a 3rd rail electrified railway is similarly dangerous to a road.

Not really; For people on foot to suffer from a third rail-electrified railway they have to go somewhere they are not allowed to be, whereas innocent pedestrians are regularly killed or injured by vehicles mounting the pavement or drivers ignoring crossing lights and signage. But this is not an argument for extending the third rail.

 

[Harpo]

And of course it is nonsense to suggest that nothing is done to address safety issues on the roads.

But it would be a nonsense to suggest that roads are never allowed to reduce safety. ‘Smart’ motorways abolishing the hard shoulder are one example.

 

[anthony263]

Im sure ive seen in a previous issue of modern railways magazine that network rail did come up with a way to improve the safety of 3rd rail to try and make it easier to get permission from the authorities to expand the network etc.

Will have to try and dig it out but if so I do think we should fill in the gaps in the network and expand other parts

 

[JamesT]

But it would be a nonsense to suggest that roads are never allowed to reduce safety. ‘Smart’ motorways abolishing the hard shoulder are one example.

Statistically, smart motorways are safer than normal ones.

 

[zwk500]

Absolutely not. The safety requirements pertaining to the third rail network as a whole should be followed.

If so, we'd be ripping out a lot of third rail

And it is a de facto ban.

No it isn't, you sound histrionic repeating this like an article of faith. The East London Line is the best example of an entire line, but also consider all the new platforms and remodellings around the third rail network that if there was a de facto ban would have had to be left unelectrified.

The requirements to be satisfied for Uckfield or Marshlink to be electrified with 3rd rail have been laid out in RSSB documentation (for those with RSSB Logins). The fact they do not justify the cost of such mitigation does not mean there is a third rail ban any more than that there is an OLE ban on the Settle & Carlisle.

 

[Magdalia]

Statistically, smart motorways are safer than normal ones.

Indeed. A huge amount of work was done on the safety implications of smart motorways by Highways England, for example see section 4 of this document:

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/936811/smart-motorway-safety-evidence-stocktake-and-action-plan.pdf

In October 2019, the Secretary of State asked the Department to carry out an evidence stocktake to gather the facts on the safety of smart motorways and make recommendations. A wide range of data has been considered and conclusions drawn on what the evidence tells us about the safety of this type of motorway. This work is set out in the Evidence Stocktake in Part A of this document. The action the Government is taking in response to this work is set out in the Action Plan in Part B of this document.

a) Modelling of Relative Risks on Motorway Types 4.5 In addition to delivering the expected benefits of increased capacity and reduced congestion, a smart motorway scheme is expected to reduce some forms of risk. This is delivered through traffic speeds becoming more consistent, the speed differential between lanes reducing, signage alerting drivers to hazards, and by providing emergency areas which are expected to be a safer place to stop than a hard shoulder. However, some other forms of risk were expected to be greater than on conventional motorway, such as the risk of a vehicle stopping in a live lane. 4.6 To understand the overall effect, prior to the introduction of DHS and ALR as types of motorway, Highways England developed a method of hazard analysis to compare the relative safety risks of the design concepts with those of a conventional motorway. This enabled Highways England to consider how the likelihood and 35 impact of around 140 hazards (and mitigations) could vary between motorways types. This assessment was further informed by a series of workshops, and evidence from additional research and testing, including a driving simulator with members of the public. 4.7 This created an assessment of risk exposure, combining the probability of risks occurring and the severity of the consequences. An example of this analysis for a conventional three lane motorway and an ALR motorway is illustrated in figure 13 below, where ‘D3M’ is a conventional 3 lane motorway, and ‘MM-ALR’ is an All-Lane Running motorway. This suggested that on smart motorways, some risks would be reduced, and some would be increased. For example, risks expected to reduce on ALR included collisions involving speeding or tailgating, while risks expected to increase included collisions involving vehicles stopped in a live lane. However, the overall expected effect was a reduction in risk. This demonstrated that while the nature of the risks varied between the motorway types, ALR was expected to reduce the overall level of risk by 20% and be as safe as, or safer than, conventional motorways.

 

[yorksrob]

If so, we'd be ripping out a lot of third rail

No it isn't, you sound histrionic repeating this like an article of faith. The East London Line is the best example of an entire line, but also consider all the new platforms and remodellings around the third rail network that if there was a de facto ban would have had to be left unelectrified.

The requirements to be satisfied for Uckfield or Marshlink to be electrified with 3rd rail have been laid out in RSSB documentation (for those with RSSB Logins). The fact they do not justify the cost of such mitigation does not mean there is a third rail ban any more than that there is an OLE ban on the Settle & Carlisle.

Well no, the fact that we're not ripping out the third rail rather demonstrates that the safety requirements pertaining to it, don't require it to be ripped out.

The East London line has been electrified by third and fourth rail since at least the 1920's.

 

[35B]

Indeed. A huge amount of work was done on the safety implications of smart motorways by Highways England, for example see section 4 of this document:

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/936811/smart-motorway-safety-evidence-stocktake-and-action-plan.pdf

An analysis that underplays the trade off between frequency and severity of risk, increasing the probability and impact of statistically rare events while reducing the overall risk level.

 

[zwk500]

The BEMU still needs partial electrification to function, in many cases we might also want to use 3rd rail to provide the "charge island" to allow them to run much further off the network.

My proposed solution would have an insulated bus bar to carry the 3rd rail current, this would have a cost commensurate with a rail. Every 100m or so this would be connected to a 100m long section of the 3rd rail via the breaker box. If a train is not present then the box does not connect the 3rd rail section to the energised bus bar, were someone/thing to bridge between the 3rd rail and a running rail no current would flow. When the box detects a train it connects the bus bar to the 3rd rail, potentially checking for a short circuit before doing so. The train passes over the energised section completing the circuit, when it leaves the circuit the box disconnects that section of 3rd rail.

Thus the only live sections of 3rd rail are under a train or recently under a train. We can therefore make an argument that we could run the 3rd rail at a much higher voltage potentially allowing much higher power transfer and much less frequent sub stations. This might potentially be a more resilient and lower cost solution than 25kv OHLE.

Since this would be used on new sections of line initially we would have 3rd rail BEMUs that were also able to operate on 750VDC 3rd rail, with modern power electronics multiple voltages are not a massive challenge.

This seems to be less beneficial than just having electrified areas and static charging bars at termini, given the lengths of the areas where third rail would be preferable instead of OLE. What's the replacement rate of your £10k a pop switches? What's the impact of the failing on the service.
Upgrading the entire 3rd rail network to higher voltage is not going to come cheap either. And how does the massive cross-section of the conductor rail affect efficiency at higher voltages?

I am struggling to see any advantages of this proposal against the 3 most viable alternatives:
- Keep 3rd rail as is (or possibly roll it back slightly), use BEMUs and provide a static charging solution at termini if required.
- Convert the 3rd rail network to 3KV DC OLE (with an eye to the French MVDC proposal).
- Convert the 3rd rail network to 25KV AC OLE.

 

[Magdalia]

Well no, the fact that we're not ripping out the third rail rather demonstrates that the safety requirements pertaining to it, don't require it to be ripped out.

Not yet.

But with a bimode 3rd rail/battery BEMU fleet the next step will be to rip out the third rail at the most high risk locations such as stations and level crossings, and run on battery power through those high risk locations.

 

[Recessio]

I don't really see why smart motorways are at all a comparison here: nobody (except maybe Sir David Serpell) is going to rip up all 3rd rail lines and replace them with motorways.

And it is a de facto ban.

As Bald Rick has so expertly explained (and it really should be bookmarked for future threads!) , there isn't a ban, it's just that new third rail has a very high barrier to surpass when compared to the alternative solutions such as a) AC OHLE, b) BEMU stock, or perhaps HM Treasury's favourite option c) do nothing

 

[yorksrob]

Not yet.

But with a bimode 3rd rail/battery BEMU fleet the next step will be to rip out the third rail at the most high risk locations such as stations and level crossings, and run on battery power through those high risk locations.

Who knows. The future is a different country.

 

[Djgr]

, there isn't a ban, it's just that new third rail has a very high barrier to surpass when compared to the alternative solutions such as a) AC OHLE, b) BEMU stock, or perhaps HM Treasury's favourite option c) do nothing

Isn't that what de facto means?

== Doublepost prevention - post automatically merged: 14 Mar 2026 ==

Indeed. A huge amount of work was done on the safety implications of smart motorways by Highways England, for example see section 4 of this document:

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/936811/smart-motorway-safety-evidence-stocktake-and-action-plan.pdf

I don't think many people will be persuaded by arguments made in justification of "smart" motorways. Perhaps the methodology employed relating to risk and safety needs revisiting?

 

[WAO]

Statistically, smart motorways are safer than normal ones.

I think that is because, in the short term, they obviously reduce traffic density (if the crawler lane is fully used). In the long term if the fourth lane runs at capacity, they will be less safe. The motoring organisations and police are clear about "smart" motorways.

In the dc case, the safety argument should be about staff and public safety, where inadvertent (not trespass) exposure is possible. The past case of children getting through a neglected fence (to an ac system) is also relevant. Expansion of dc beyond its SE limits is technically not reasonable and I would argue that Merseyrail should have the CVL solution of partial ac and batteries as should Pirbright Jn to Weymouth.

WAO

 

[zwk500]

Isn't that what de facto means?

No, a de facto ban would mean it isn't permitted in practice at all.

It *is* permitted in practice, it's just that the lines where some people want to really see it don't justify the mitigation measures. That's not the same as a ban. As I mentioned a couple of posts up, nobody would suggest that OLE is 'banned' from the Settle & Carlisle line.

 

[yorksrob]

No, a de facto ban would mean it isn't permitted in practice at all.

It *is* permitted in practice, it's just that the lines where some people want to really see it don't justify the mitigation measures. That's not the same as a ban. As I mentioned a couple of posts up, nobody would suggest that OLE is 'banned' from the Settle & Carlisle line.

There's no economic likelihood of OLE on the S&C, however there's no safety assumption against it either, so that's a non-comparison.

 

[zwk500]

There's no economic likelihood of OLE on the S&C, however there's no safety assumption against it either, so that's a non-comparison.

Are you suggesting OLE doesn't require safety mitigation?

 

[tbtc]

It's amazing that, after over two hundred posts on a thread titled 'Should the third rail ban be lifted?', there's still no actual evidence that there is any actual 'ban

(clearly there are examples of modern extensions to Third Rail, when that has been the most appropriate option, as evidenced on here)

What we have instead are:

1. A need for future schemes to be the better of the various options (AC, DC, Battery etc - plus doing nothing). This doesn't seem controversial to me. Of course when you are spending millions of public money there should be a demonstration that the chosen option is the best one. Unfortunately for Third Rail that means a lot more fencing/ closing crossings etc given the additional risks of people on electrified track, so it's not particularly competitive

2. A huge and seemingly growing wishlist of lines near existing DC routes which some people seem convinced would magically be electrified if only there wasn't this 'ban'. Which no doubt sounds like a nice story to tell fellow believers but obviously ignores the fact that there are similarly loads of unelectrified lines just beyond OHLE - nobody is claiming that the fact there's no wires at Buxton/ Harrogate/ Morecambe (or Crewe to Chester, Durham Coast, that frustratingly short spur at Morpeth that BR didn't do... Or Oxford/ Bristol/ Lincoln/ Swansea) etc is evidence of a 'ban' on AC electrification. It's only relatively recently that some low hanging fruit like Stoke - Crewe was finally done

3. The idea that 'oh BR would absolutely have done all of these things', which combines the fact that BR were much better at conceptual plans for the future than they were at actually doing the day job , the fact that BR were still closing lines/stations into the eighties and nineties, the fact that BR were downgrading lots of things (bus shelters replacing proper waiting rooms, two coach Sprinters replacing loco hauled, 153s then replacing two coach Sprinters...) - all the Artists Impressions of Stuff BR Wanted To Do If It Had An Unlimited Budget were all well and good but that's all most of them ever were. Pretty pictures.

4. Some slightly fanatical people who seem a lot more focused on certain 'solutions' than they are on examining what the problem actually is. Any type of electrification is only ever a means to an end, in the same way that reopening some rural route is only a means to an end. But if you approach every discussion with the attitude that 'here's my preferred solution, regardless of what the actual problem is' then it starts to look like an obsession. When all your have is a hammer, every problem looks like a nail. And when all you have is a fanatical devotion to your certain cause, every issue magically looks like a chance to reheat the same old tired ideas

 

[yorksrob]

Are you suggesting OLE doesn't require safety mitigation?

No, I'm saying that there's no safety assumption against it.

 

[Djgr]

No, a de facto ban would mean it isn't permitted in practice at all.

It *is* permitted in practice, it's just that the lines where some people want to really see it don't justify the mitigation measures. That's not the same as a ban. As I mentioned a couple of posts up, nobody would suggest that OLE is 'banned' from the Settle & Carlisle line.

In practice it isn't permitted, because of the onerous and inappropiate requirements.

 

[zwk500]

No, I'm saying that there's no safety assumption against it.

The comparison still holds though - An OLE proposal must justify the costs required to implement the mitigation measures to make the system safe, just as a third rail proposal must.

The fact the third rail requires much more mitigation work because it is far more dangerous does not mean it is banned.

In practice it isn't permitted,

So the Bricklayers Arms Flyover, Gatwick Airport Platform 7, new Rochester Station, Gillingham extra platform, and all the other enhancements were done without permission? They all involved new third rail.

because of the onerous and inappropiate requirements.

Onerous yes. Inappropriate is subjective though, and is not the view of those who's legal responsibility it is to set the requirements.

 

[Bald Rick]

And it is a de facto ban.

“De Facto” means (almost literally) that it is a matter of fact or certainty. As we know, there are plenty of recent examples of new third rail installations, so, clearly, there isn’t a ban.

Well no, the fact that we're not ripping out the third rail rather demonstrates that the safety requirements pertaining to it, don't require it to be ripped out.

Correct. But don’t confuse ‘not ripping out’ with ‘therefore new installations are as safe as is reasonably practical‘. They are two very different things. I’ll explain this another time, but the short version is that, very clearly, it is not reasonably practical to buy hundreds of new BEMUs (or AC EMUs) when a small part of the existing DC network needs replacement or upgrading. Because we are not replacing the whole network in one go (or even part of it) or the DC whole fleet*, then clearly there will not be a safety driven case for it. There would be far better places to spend that money for safety purposes, such as closing lots of level crossings.

* when all the new Southeastern metro fleet turns up with batteries, as it surely will, there will then be some interesting potential on areas where only those units run. However with the exception of the end of the Hayes branch I can’t think of many parts of the network where only those units will run.

The East London line has been electrified by third and fourth rail since at least the 1920's.

Not the bit between Dalston and just north of the new Shoreditch, that didn’t exist for 30 years. And the bit from just north of the new Shoreditch to old Shoreditch, and the bit over the floyover st New Cross Gate, and all the depot at New Cross Gate, and the new stretch around Surrey Cnaal Road… none of that existed at all.

 

[yorksrob]

“De Facto” means (almost literally) that it is a matter of fact or certainty. As we know, there are plenty of recent examples of new third rail installations, so, clearly, there isn’t a ban.




Correct. But don’t confuse ‘not ripping out’ with ‘therefore new installations are as safe as is reasonably practical‘. They are two very different things. I’ll explain this another time, but the short version is that, very clearly, it is not reasonably practical to buy hundreds of new BEMUs (or AC EMUs) when a small part of the existing DC network needs replacement or upgrading. Because we are not replacing the whole network in one go (or even part of it) or the DC whole fleet*, then clearly there will not be a safety driven case for it. There would be far better places to spend that money for safety purposes, such as closing lots of level crossings.

* when all the new Southeastern metro fleet turns up with batteries, as it surely will, there will then be some interesting potential on areas where only those units run. However with the exception of the end of the Hayes branch I can’t think of many parts of the network where only those units will run.




Not the bit between Dalston and just north of the new Shoreditch, that didn’t exist for 30 years. And the bit from just north of the new Shoreditch to old Shoreditch, and the bit over the floyover st New Cross Gate, and all the depot at New Cross Gate, and the new stretch around Surrey Cnaal Road… none of that existed at all.

I must admit, I'd forgotten about the old Broad Street line.

However, if they can countenance new third rail in crowded urban areas like North London where the risk of trespass must be so much higher, the inability to do the infills in rural Kent and Sussex seems even more non-sensical than I'd previously thought.

 

[35B]

I must admit, I'd forgotten about the old Broad Street line.

However, if they can countenance new third rail in crowded urban areas like North London where the risk of trespass must be so much higher, the inability to do the infills in rural Kent and Sussex seems even more non-sensical than I'd previously thought.

Given that the line between Dalston and Shoreditch is on viaduct or closed embankment, it’s less ridiculous than you suggest. It would be more interesting to understand the mitigations for the section between Surrey Quays and South Bermondsey, which is much less enclosed.

I do note that the panacea for avoiding infrastructure investment (batteries) was much less available at that time.

 

[zwk500]

However, if they can countenance new third rail in crowded urban areas like North London where the risk of trespass must be so much higher, the inability to do the infills in rural Kent and Sussex seems even more non-sensical than I'd previously thought.

Look at the lines. They were justified because access was heavily restricted by the vertical separation from street/ground level.

It would be more interesting to understand the mitigations for the section between Surrey Quays and South Bermondsey, which is much less enclosed.

Have a look on google street view of the level of security fencing provided. And that's with a fair proportion of the line on an embankment or viaduct, and there are no level crossings (only 1 maintenance access point, which can only be reached from the Depot).

I do note that the panacea for avoiding infrastructure investment (batteries) was much less available at that time.

This is also relevant - the impact of not fitting third rail to that section would have been to require either battery traction (which wasn't widely available as you note), a very brief section of OLE, or a diesel/hydrogen option.

It is very possible that if BEMUs were not being procured for Merseyrail, the very similar section to Headbolt Lane would have also been electrified. (IMHO, it should have been anyway).

 

[yorksrob]

Look at the lines. They were justified because access was heavily restricted by the vertical separation from street/ground level.

Urban viaducts are notoriously tempting for budding Banksies. The trespass risk in North London must be through the roof.

The Romney Marsh, less so.

 

[renegademaster]

Miliband reveals plans that could mean nuclear power plants built near homes

Changes to regulation to speed up development could also make it easier to build on sensitive nature sites

www.theguardian.com

Ed Miliband has unveiled plans that could make it easier to build nuclear power plants closer to homes and on sensitive nature sites, as he attempts to speed up the development of energy infrastructure.

The current government in other areas can sometimes recognizes the existing regulations as not an unquestionable good and can accept raising risk slightly is ok if it's for the greater good

 

[Harpo]

Urban viaducts are notoriously tempting for budding Banksies. The trespass risk in North London must be through the roof.

ORR was presumably satisfied that the scheme’s safety controls made the resultant risk profile ALARP.

The Romney Marsh, less so.

But it will be very expensive to make it’s overall risk level match TfL’s recent additions.

 

[Bald Rick]

However, if they can countenance new third rail in crowded urban areas like North London where the risk of trespass must be so much higher, the inability to do the infills in rural Kent and Sussex seems even more non-sensical than I'd previously thought.

Urban viaducts are notoriously tempting for budding Banksies. The trespass risk in North London must be through the roof.

I can advise from experience that trespass on the ELL from Dalston to Shoreditch is near non-existent. It is almost impsosible to access the tracks except from the platform, and the stations have protective measures including full CCTV, staffing first till last train, and security guards at certain times.

Of course, what has changed since the risk assessments for the ELL were done (back in the mid-2000s) is that the cost and practicality of alternatives - particularly battery trains - has changed significantly. It is reasonable to assume that if the ELL was in development now for construction in a few years’ time with a completely new fleet, then the assessment would most likely point towards a battery fleet with no con rail north of Whitechapel.

 

[Recessio]

Miliband reveals plans that could mean nuclear power plants built near homes

Changes to regulation to speed up development could also make it easier to build on sensitive nature sites

www.theguardian.com

The current government in other areas can sometimes recognizes the existing regulations as not an unquestionable good and can accept raising risk slightly is ok if it's for the greater good

With the danger of veering off topic, the risk level (which is a measure of both the impact of the danger, and the probability/likelihood/rate of occurrence of the danger) of living next to a nuclear plant is much smaller than the risk of someone being electrocuted on a conductor rail, for which statistics have been provided earlier in this thread