ERA Report on Harmonisation of EU Electrification Standards

[zwk500]

I've posted in here because it's relevant to the UK on 2 fronts: the UK's connection to the European Rail network and the UK's own 'should we convert third rail' conundrum. But mods, please move to International if more apppropriate!

The ERA have published a report looking at the case for re-electrifying Europe from the current four permitted standards (1.5kV DC, 3kV DC, 15kV AC, 25kV AC) to 3, 2, or even standardising on 25KV AC. This is with the objective of simplifying cross-border trains.

Their conclusion was largely that for normal traffic it wasn't worth it, as Multi-Mode locos and trainsets are now the standard and this has meant the costs of conversion usually outweigh the efficiency gains. They acknowlege there are individual cases were re-electrification to remove incompatible systems is desirable, but also note the emergence of battery as an option.
For high-speed rail, however, their conclusion is that re-electrifying DC sections of classic lines used by High-speed trainsets (or bypassing them) to simplify the costs of these vehicles may be worthwhile.

Full report available here:https://www.era.europa.eu/content/era-study-harmonisation-electrification-systems-0

The study identified that there would be substantial benefits to having a single electrification system. However, there is no straightforward migration scenario or clear economic case for re-electrification to a target system, thus confirming the conclusions of the ERA 2009 impact assessment on the TSI ENE.
In fact, the study found that interoperability barriers created by differences in voltage and frequency are mitigated by multi-system traction vehicles. The fleet analysis showed that the share of multi-system locomotives is increasing, with four-system locomotives being de facto the standard.
This report therefore does not recommend a complete Europe-wide, large-scale re-electrification, but would support reflection on whether there are specific sections of the network that could be re-electrified to benefit the railway system (e.g. where AC lines are interrupted by relatively small DC sections). Such reflection should occur in broad consultation with infrastructure managers and operators, on a case-by-case basis, with a longterm perspective.
A lack of electrification can increase the cost of cross-border operations due to locomotive changes and associated delays. Thus, non-electrified border crossings can be impediments to international traffic. That said, this study does not suggest that all border crossings and lines need to be electrified. Other solutions, such as battery trains, may be more cost-efficient, including for cross-border areas where lines with multiple voltages exist. Additionally, there are several border crossings where the voltage of one country is extended to the first station after a border, enabling passengers to cross borders without necessitating the usage of multi-system rolling stock. This solution could reduce operating costs while still allowing cross-border operations, with a switchover happening at the border station.

Point 3: high-speed rail
Given that it seems clear that multiple electrification systems will endure for a long time, the need for multi-system traction vehicles has to be acknowledged.
In the case of locomotives, it was observed that market dynamics have de facto made multi-system solutions the standard, despite the additional complexity and costs compared with mono-system options. Economies of scale and innovation have reduced the costs sufficiently to make multi-system locomotives commercially attractive. In the case of trainsets, however, no large-scale deployment of multi-system vehicles could be observed.
The recently published high-speed rail plan emphasises the need for interoperable rail infrastructure and rolling stock. The argument is that the harmonisation of requirements reduces costs, improves production rates and ensures greater areas of use.
Since 2014, the TSI ENE has mandated that new lines with speeds greater than 250 km/h must be supplied with one of the AC systems. The consequence is that predominantly DC networks also integrate AC lines. Yet it is common for parts of the high-speed line (e.g. in stations or urban areas) to be DC lines, thus requiring high-speed rolling stock to run on DC.
From a European system development perspective, it is important to determine whether it is more beneficial to focus on the development of multi-system high-speed trainsets to handle different voltages, similar to what has happened for locomotives, or to invest in the re-electrification of selected DC sections along AC high-speed lines or in bypasses to separate AC and DC networks.

This largely confirms the previous discussions about re-electrifying the Southern Region not being worthwhile, although noting that the Third Rail network has the added factor of the increased safety risk from the ground-level conductor.

 

[GraphitProjekt]

They don't seem to mention that any of the standard voltages are particularly good or bad for freight movements. I think this is especially an issue for our 750V network mainly because the Channel Tunnel is there and HS1 is not really suitable. I think the only lines that should be seriously considered for 25kV conversion are those for major freight movements towards Southampton and Dollands Moor. It might be easier in the latter case to do a freight only mainline like was done in the Netherlands, or at least a few choice bypasses. For Southampton it's arguably a bit easier because you're pretty much at the edge of the DC network anyway, so the number of interfaces isn't too high.

The one thing that I think is most exciting about theoretically converting DC to AC, is that you might be able to re-use the 33kV distribution network since the trial at Doncaster shows that you can run a fairly powerful 25kV supply off of a 3phase 33kV source. This would give a very different architecture of 25kV feeding compared to the way it's normally done

 

[zwk500]

They don't seem to mention that any of the standard voltages are particularly good or bad for freight movements.

Yes, it's interesting to see there was almost zero discussion of the technical advantages and disadvantages of the different systems, although that work was presumably done as part of the initial TSI work that led to the adoption of 4 standards, but with a presumption of 25kV.

I think this is especially an issue for our 750V network mainly because the Channel Tunnel is there and HS1 is not really suitable. I think the only lines that should be seriously considered for 25kV conversion are those for major freight movements towards Southampton and Dollands Moor.

It is difficult to say there's any major freight movements towards Dollands Moor. And the big issue for access to Dollands Moor is the requirement to use the West London Line and Clapham Junction (if you're not using HS1 all the way to Barking). Conversion of any of the 3 nominated Channel Tunnel-Wembley routes to 25KV would create substantial numerous interfaces with third rail.

It might be easier in the latter case to do a freight only mainline like was done in the Netherlands, or at least a few choice bypasses.

I don't see where you could reasonably build a freight only mainline between Dollands Moor and the south WCML. Beefing up the 750v network would presumably be easier, especially if using a battery loco to even out power draw?

For Southampton it's arguably a bit easier because you're pretty much at the edge of the DC network anyway, so the number of interfaces isn't too high.

Southampton is interesting for the possibilities battery/25KV locos offer, which probably negate any need to interface directly with the DC network.

The one thing that I think is most exciting about theoretically converting DC to AC, is that you might be able to re-use the 33kV distribution network since the trial at Doncaster shows that you can run a fairly powerful 25kV supply off of a 3phase 33kV source. This would give a very different architecture of 25kV feeding compared to the way it's normally done

Is this cheaper or more efficient than traditional feeding methods, or is the advantage that you don't need to wait for a brand new National Grid connection?

 

[GraphitProjekt]

It is difficult to say there's any major freight movements towards Dollands Moor. And the big issue for access to Dollands Moor is the requirement to use the West London Line and Clapham Junction (if you're not using HS1 all the way to Barking). Conversion of any of the 3 nominated Channel Tunnel-Wembley routes to 25KV would create substantial numerous interfaces with third rail.

Out of interest, what are the main routes to Wembley?

I don't see where you could reasonably build a freight only mainline between Dollands Moor and the south WCML.

I was thinking more along the lines of the old Electric Spine concept that joins up with the more achievable Southampton re-electrification. So freight for Dollands Moor, comes through EWR from Bletchley through to Didcot and Reading where conversions and bypasses start appear going the long way round to Dollands Moor. Extreme fantasy stuff of course and in the medium term, trying to use HS1 is a better option, but it will always be limited by night-time only use and the expensive upkeep needed to keep HSR tolerances while the track is bashed around by hundreds of tonnes of freight.

Beefing up the 750v network would presumably be easier, especially if using a battery loco to even out power draw?

It would be interesting to know what the limiting factor is with 3rd rail freight. If it's just expensive then that does mean it's doable. One thing that comes to mind is the potential issue that such high starting currents might require the fault current to be set so high that substations will stop being able to detect most faults. But indeed on-board batteries could help with that.

However, worth noting that some routes were specifically upgraded to enable class 92 and class 373 movements, which apparently did used to happen not that long ago So 3rd rail freight might actually be fine, it just never happens because it's expensive and so not worth it for freight companies who have to make money on every movement. But if as a nation we got really serious about utilising Dollands Moor and the Channel Tunnel, maybe it could just be done with a few PSUs. (Indeed, DC PSUs are somewhat more straightforward than 25kV PSUs).

Southampton is interesting for the possibilities battery/25KV locos offer, which probably negate any need to interface directly with the DC network.

But for this route, the number of interfaces is a more reasonable 4. Basingstoke (east of Reading Ln jnc), Eastleigh (for the Fareham line), St Denys, and the last maybe quite a way west of Southampton maybe near Ashurst New Forest. There is also a question about whether you want interfaces in motion or within stations where all trains would stop, the latter I think makes placement a lot more complex and probably requires more stretches of dual electrification, but the former may or may run into an issue with line speeds and how fast you can reasonably drop shoes.

Is this cheaper or more efficient than traditional feeding methods, or is the advantage that you don't need to wait for a brand new National Grid connection?

Not especially cheaper at least right now while it's a trial with a few examples. But it's definitely more efficient with the possibility of dual end parallelised feeding with far fewer neutral sections needed, this also makes it much more scalable to upgrading to new demand compared to normal 25kV. I would imagine it's a lot quicker to secure SFC connections as increasingly little 132kV infra is suitable for single phase 25kV anymore (depends on the area), so then you have to bend yourself around where transmission lines happen to cross your line and then live with the decades it takes to follow through such connections. If you can connect like a normal industrial customer, everything's a lot quicker and perhaps easier to agree and there are many more options available to you allowing you to install more rational feeding arrangements. GWML is in some ways an example of having to fit everything around where suitable 400kV infrastructure is available rather than what might be more optimal for the railway. I think if your line is too far away from any transmission infrastructure, SFCs be the only option.

 

[zwk500]

Out of interest, what are the main routes to Wembley?

From the Sectional Appendix:
CTRL 1 Wembley-Latchmere Junction-Nunhead/Herne Hill-Maidstone East-Dollands Moor
CTRL 2 Wembley-Latchmere Junction-Orpington/Sevenoaks or Swanley/Otford/Sevenoaks-Dollands Moor
CTRL 3 Wembley-Latchmere Junction-Redhill-Tonbridge-Dollands Moor

They're mentioned in relation to being exempt from the Two-Thirds Rule, but believe there are also restrictions around maintenance access having to leave at least one open, and they received gauge enhancements, loop upgrades, and signalling work in advance of the channel tunnel.

I'll have a proper read of your other points later.

 

[yorksrob]

I've posted in here because it's relevant to the UK on 2 fronts: the UK's connection to the European Rail network and the UK's own 'should we convert third rail' conundrum. But mods, please move to International if more apppropriate!

The ERA have published a report looking at the case for re-electrifying Europe from the current four permitted standards (1.5kV DC, 3kV DC, 15kV AC, 25kV AC) to 3, 2, or even standardising on 25KV AC. This is with the objective of simplifying cross-border trains.

Their conclusion was largely that for normal traffic it wasn't worth it, as Multi-Mode locos and trainsets are now the standard and this has meant the costs of conversion usually outweigh the efficiency gains. They acknowlege there are individual cases were re-electrification to remove incompatible systems is desirable, but also note the emergence of battery as an option.
For high-speed rail, however, their conclusion is that re-electrifying DC sections of classic lines used by High-speed trainsets (or bypassing them) to simplify the costs of these vehicles may be worthwhile.

Full report available here:https://www.era.europa.eu/content/era-study-harmonisation-electrification-systems-0




This largely confirms the previous discussions about re-electrifying the Southern Region not being worthwhile, although noting that the Third Rail network has the added factor of the increased safety risk from the ground-level conductor.

True, and we have already abided by any such suggestion by building HS1

 

[edwin_m]

Is this [using the 33kV feeder network if converting 750V routes to 25kV] cheaper or more efficient than traditional feeding methods, or is the advantage that you don't need to wait for a brand new National Grid connection?

Probably the latter. The 33kV network feeds all the substations along the line without each one needing its own Grid/DNO connection. 25kV lines don't have similar lineside feeders (except in a few special cases) so the easiest option would be abandoning the 33kV distribution and providing a 25kV feeder station everywhere that network was fed from the Grid. That Grid feed would come in at a much higher voltage anyway, so the ability of static converters to draw at lower voltages isn't relevant, but using them instead of a conventional feeder would avoid introducing phase imbalance (DC feeders draw from all three phases equally).

It would be interesting to know what the limiting factor is with 3rd rail freight. If it's just expensive then that does mean it's doable. One thing that comes to mind is the potential issue that such high starting currents might require the fault current to be set so high that substations will stop being able to detect most faults. But indeed on-board batteries could help with that.

Likely reasons would include the fault current issue and the related issue of voltage drop, which can be addressed by adding more substations but it's a diminishing return. More current for more time also increases the heat dissipated even if it stays below substation tripping levels, which along with a warm climate might require quite a few to be updated. International locomotive designs that might be shrunk to fit in the UK loading gauge might not be able to accept such a low supply voltage considering they were designed with supplies of 1.5kV upwards in mind, so the traction packages might need re-designing too.

 

[BRX]

specifically upgraded to enable class 92 and class 373 movements, which apparently did used to happen not that long ago So 3rd rail freight might actually be fine

92s were routine between dollands moor and wembley until about 10+ years ago. My impression is that DB just decided it wasn't worth keeping their 3rd rail capability going - and they were then confined to the tunnel and HS1. They were simply replaced by 66s on the "classic" route to wembley. This was in the context of dwindling tunnel traffic but I think also more and more of DBs 92s having problems so they became concentrated where they were essential - moving traffic through the tunnel itself.
GBRf have two 92s based at DM that can still use 3rd rail and they are still very occasionally used for "real" freight but most of the time if they are on 3rd rail they are running light engine.

So yes it's perfectly possible but perhaps only with some more attentive maintenance of the locos, which are well known as being rather complex.

I'm always wondering if a new 3rd rail freight loco might appear some day.

 

[GraphitProjekt]

I'm always wondering if a new 3rd rail freight loco might appear some day.

Good to hear some DC capability is still there with 92s. Yeah ultimately we'd need a new loco type. But it'd have to be a centrally administered project because I don't think any freight company will be able to invest in a new fleet with limited capabilities especially when the freight sector is largely unsupported. Maybe some variation of the class 99 with shoegear could do it? But again, some kind of GBR investment and administration has to happen to get the companies to actually run on electric and make the any additional infrastructure investments beyond the PSUs done with Channel Tunnel enabling works.

Probably the latter. The 33kV network feeds all the substations along the line without each one needing its own Grid/DNO connection. 25kV lines don't have similar lineside feeders (except in a few special cases) so the easiest option would be abandoning the 33kV distribution and providing a 25kV feeder station everywhere that network was fed from the Grid.

Well the current DC traction substations typically supply signalling power as well, so the easiest move is to retain them for that which would go along with the HV feeders. There is something of an added resiliency benefit as well since you can feed 25kV from a point that is itself fed from elsewhere. Also you get the opportunity to add more 25kV connections (if needed, obviously never at All or even a majority of 750V sites) to boost the supply in a manner much more like how 15kV AC railways are run, where individual supply points are a less powerful (and therefore substantial/expensive) but greater in number because they can all operate in parallel

That Grid feed would come in at a much higher voltage anyway,

Well in the majority of cases, the railway 33kV feeders merely draw off of 33kV switchboards at 132kV substations, not off the 132kV side itself. There are a small number of exceptions to this (which I made a thread about). Indeed each 25kV supply point would probably be a lot cheaper drawing off 33kV and require little alteration to the existing power network - but they would still be able to provide a lot more power because of the aforementioned double-ended/parallel feeding arrangement which normal 25kV can't do.

 

[Sir Felix Pole]

One suggestion in the past has been to route Freighliner &c services to Southampton (Millbrook) via Laverstock (Salisbury) with 25kV AC electrification Reading - Basingstoke - Salisbury- Soton. This reduces interfaces with the third-rail, frees up capacity via Eastleigh, and significantly improves passenger services. It would also go a long way to solving the 'Exeter' problem, giving battery trains nore opportunity to charge.

 

[edwin_m]

Well the current DC traction substations typically supply signalling power as well, so the easiest move is to retain them [33kV feeders] for that which would go along with the HV feeders. There is something of an added resiliency benefit as well since you can feed 25kV from a point that is itself fed from elsewhere. Also you get the opportunity to add more 25kV connections (if needed, obviously never at All or even a majority of 750V sites) to boost the supply in a manner much more like how 15kV AC railways are run, where individual supply points are a less powerful (and therefore substantial/expensive) but greater in number because they can all operate in parallel

Seems a lot of kit to maintain just to supply the signalling which would be a tiny fraction of the traction load this network was designed to support. I guess it could be retained initially then progressively abandoned as other arrangements were made.

Well in the majority of cases, the railway 33kV feeders merely draw off of 33kV switchboards at 132kV substations, not off the 132kV side itself. There are a small number of exceptions to this (which I made a thread about). Indeed each 25kV supply point would probably be a lot cheaper drawing off 33kV and require little alteration to the existing power network - but they would still be able to provide a lot more power because of the aforementioned double-ended/parallel feeding arrangement which normal 25kV can't do.

Good point. Perhaps different arrangements would be made case by case depending on things like the age of the existing equipment, how much extra power was needed, and whether converting DC system Grid feeds into AC ones would result in a workable distribution of feeders around the network without having to keep the 33kV cables.

 

[GraphitProjekt]

One suggestion in the past has been to route Freighliner &c services to Southampton (Millbrook) via Laverstock (Salisbury) with 25kV AC electrification Reading - Basingstoke - Salisbury- Soton. This reduces interfaces with the third-rail, frees up capacity via Eastleigh, and significantly improves passenger services. It would also go a long way to solving the 'Exeter' problem, giving battery trains nore opportunity to charge.

Is this better with regards to capacity on SWML? I would support that as a potentially better solution to re-electrifying the main route as you get some more passenger wiring as well. Is that okay for loading gauge?
What would happen around the edges though? Dual voltage railway between Reading Line jnc to Worting Jnc?

 

[edwin_m]

Is this better with regards to capacity on SWML? I would support that as a potentially better solution to re-electrifying the main route as you get some more passenger wiring as well. Is that okay for loading gauge?
What would happen around the edges though? Dual voltage railway between Reading Line jnc to Worting Jnc?

According to the map in this link, Southampton via Laverstock is already cleared W12, which is actually better than via Winchester. Probably not 25kV clearances though.

Gauging or Gouging? It’s all in a few millimetres - Rail Engineer

Listen to this article Early railways were built for freight; passenger traffic was a slightly unexpected bonus. Engineers had to be sure that the goods on trains would fit through tunnels and under bridges and the concept of the load gauge was born. It was an inverted semicircle of metal hung...

www.railengineer.co.uk

There was an RSSB project about 15 years ago that looked at converting the third rail network, which proposed starting with Southampton to Reading. As well as the freight use, the EMUs using it were built with dual voltage operation in mind, and could relatively easily be retrofitted and change system near Basingstoke. However, if that ever was a realistic idea, cancellation of Reading to Oxford and the "Electric Spine" project on to the Midlands means any electric freight could only go towards South Wales.

 

[zwk500]

What would happen around the edges though? Dual voltage railway between Reading Line jnc to Worting Jnc?

Battery, since it'll be needed for the container terminals anyway.
The length of 3rd rail traversed by a train running Reading-Basingstoke-Laverstock-Redbridge-Southmapton is a handful of miles.

 

[Mordac]

Any word about bridge clearances?

 

[zwk500]

Any word about bridge clearances?

In what respect?

 

[Mordac]

In what respect?

The change in EU electrification standards were reportedly the cause of a lot of cost increases during the 10s' electrification projects. is there any further change?

 

[zwk500]

The change in EU electrification standards were reportedly the cause of a lot of cost increases during the 10s' electrification projects. is there any further change?

This report doesn't appear to consider that level of detail, it uses a generalised cost assumption.

== Doublepost prevention - post automatically merged: 2 Jun 2026 ==

I was thinking more along the lines of the old Electric Spine concept that joins up with the more achievable Southampton re-electrification. So freight for Dollands Moor, comes through EWR from Bletchley through to Didcot and Reading where conversions and bypasses start appear going the long way round to Dollands Moor.

Freight to Dollands Moor via Bletchley and Reading is going a very, very long way round, especially as there's no curve at Redhill for cross-country traffic and no reasonable place to put one. 4-tracking the WLL would be a more realistic goal, and that's saying something!

It would be interesting to know what the limiting factor is with 3rd rail freight. If it's just expensive then that does mean it's doable. One thing that comes to mind is the potential issue that such high starting currents might require the fault current to be set so high that substations will stop being able to detect most faults. But indeed on-board batteries could help with that.

However, worth noting that some routes were specifically upgraded to enable class 92 and class 373 movements, which apparently did used to happen not that long ago So 3rd rail freight might actually be fine, it just never happens because it's expensive and so not worth it for freight companies who have to make money on every movement. But if as a nation we got really serious about utilising Dollands Moor and the Channel Tunnel, maybe it could just be done with a few PSUs. (Indeed, DC PSUs are somewhat more straightforward than 25kV PSUs).

I understand that the 92s can cause issues with signalling due to their huge power draw (specc'd for a full redundancy in the Tunnel, of course). It's possible a newer 3rd rail loco wouldn't require such a high power output so would have less issues.

But for this route, the number of interfaces is a more reasonable 4. Basingstoke (east of Reading Ln jnc), Eastleigh (for the Fareham line), St Denys, and the last maybe quite a way west of Southampton maybe near Ashurst New Forest. There is also a question about whether you want interfaces in motion or within stations where all trains would stop, the latter I think makes placement a lot more complex and probably requires more stretches of dual electrification, but the former may or may run into an issue with line speeds and how fast you can reasonably drop shoes.

Operationally the most resilient option is to put a traction change at a station stop. Basingstoke/Hook and St Denys would be simple enough. Eastleigh could be negated by extending towards Fareham or using batteries to bridge a gap between systems. However I would not have a 4th interface, but rather extend OLE down to Poole and then use batteries to Weymouth, with local charging in Weymouth station and Jubilee Sidings. The Lymington branch might need a certain amount of OLE as well to give sufficient time for charging.

Not especially cheaper at least right now while it's a trial with a few examples. But it's definitely more efficient with the possibility of dual end parallelised feeding with far fewer neutral sections needed, this also makes it much more scalable to upgrading to new demand compared to normal 25kV. I would imagine it's a lot quicker to secure SFC connections as increasingly little 132kV infra is suitable for single phase 25kV anymore (depends on the area), so then you have to bend yourself around where transmission lines happen to cross your line and then live with the decades it takes to follow through such connections. If you can connect like a normal industrial customer, everything's a lot quicker and perhaps easier to agree and there are many more options available to you allowing you to install more rational feeding arrangements. GWML is in some ways an example of having to fit everything around where suitable 400kV infrastructure is available rather than what might be more optimal for the railway. I think if your line is too far away from any transmission infrastructure, SFCs be the only option.

Interesting!