Electrifying Mainlines for Battery Train operation (using Class 897 as an example

[Nottingham59]

This thread is to discuss the electrification needed on mainlines for 125-mph capable battery trains.

To start things off, these are my thoughts for main lines in my part of the world (i.e. the MML and the Cross Country route)

BEMU SPECIFICATION
It seems to me that the 10-car Class 897 trimode, or similar, as ordered by LNER would be a good candidate for modelling the OHLE needed for batterifcation of mainlines. The 897 has a claimed range of 120 miles on batteries, after which the diesel engines would need to be run. We don't have the full 897 specification, so I'm going to assume the following:

• Maximum range on battery alone: 120 miles
• Usual operating range (80% to 20% charge): 60% x 120 = 72 miles
• Time to recharge (20% to 80%): 18 minutes (This is equivalent to a "2C" recharge rate)
• Time to recharge (80% to 90%): 6 minutes (1C rate)
• Time to recharge (90% to 100%): 12 minutes (0.5 C)
• Maximum power draw: 6MW (with supplementary pantographs for use only when stationary or moving slowly)

INITIAL OVERHEAD ELECTRIFICATION
To start off, I'm assuming that we wire all sections of routes which carry 7 or more trains per hour in each direction. According to this thread, that means Nottingham-Trent East (7tph); Dore-Sheffield (8tph), Sheffield-Meadowhall (11tph). Also Bristol TM to Parkway (~7tph)

Trivia: What sections of unelectrified line carry the most trains per hour?

Considering lines that are not yet electrified, what sections carry the most trains per hour (tph)? At this stage I'm only considering timetabled passenger services, and not including freight or ECS/NR movements. Or station throats. I can think of the following: Dore - Sheffield (8 tph, I...

www.railforums.co.uk

Other wiring could be added as needed, but these would not be a pre-requisite for running 897s with their diesel capability.

NOTTINGHAM-LONDON
With several hours under the wires, all Down trains will pass Wigston with 100% charge. 25 miles to Trent East will take that down to ~80%, followed by at least 18 minutes to recharge (6 transit; 6 dwell; 6 transit), taking them back up to 100% again by Trent East on the way back south.

SHEFFIELD-LONDON
Wigston-Dore is 65 miles, easily within the operating range of the 897. Dore-Sheffield is ~5 minutes each way and most London trains have at least 15 minutes dwell or more at Sheffield. State of charge: Wigston 100%, Dore 45%, aSHF 62%, dSHF 90%, Dore 97%, Wigston 42%. Easily done without diesel.

EDINBURGH-PLYMOUTH
Southbound Cross Country services will be 100% charged at Moorthorpe, and will still have 87% charge by Meadowhall, restored to ~97% in the 12 minutes from Meadowhall to Dore. 76 miles to Birmingham (36%). 6 mins dwell at New St (56%) plus 12 mins to Kings Norton (85%); 75 miles to Westerleigh Jn (22%). 20 minutes under the wires will get back over 80% leaving Temple Meads, which is enough to get to Tiverton before having to start the diesels. Then run on Diesel and Battery to Plymouth and back to Temple Meads.

CARDIFF-NOTTINGHAM
Trains will leave Severn Tunnel Junction with 100% charge. 73 miles to Bromsgrove (39%); 6 mins to Longbridge (59%); 2 miles to Kings Norton (57%); 12 min to New St (77%), plus 6 min dwell (82%); 50 miles to Trent East (40%). At least 18 mins to recharge at Nottingham (~90+) before passing Trent East heading back south

CONCLUSION
So it seems to me that with trimodes similar to Class 897 you can batterify the entire MML mainline and Cross Country route as far as Tiverton with just 10 route miles of OHLE at Meadowhall-Dore, 7 miles Trent-Nottingham and 5 miles between Briston Parkway and Temple Meads.

The key characteristic for the BEMU is a fast recharge rate under the wires. This dramatically shortens the length of wiring needed at each island of electrification.

Is this analysis realistic? What are your thoughts?
If you don't think it would work, what size battery would be needed?

 

[Zomboid]

Batteries depend on the chemistry used, the max/ min state of charge and charging rate, all of which will be balanced against the desired lifespan of the batteries. Ambient temperature impacts the ability to charge and discharge efficiently, and I've no idea what the 897 claims are based on.

There's a lot of assumptions in your analysis, and it's likely that initially at least a fairly conservative approach will be taken, and if similar chemistry to most cars is used they'll usually operate in the 20-80% state of charge range.

Also, the what bits to electrify debate will also depend on external factors, such as where a supply is available, and the required resilience of the supply. 897s having diesel means lower resilience would be acceptable for those, but that's not where things really should go in the medium term - getting rid of the diesel entirely should be the target.

 

[hwl]

This thread is to discuss the electrification needed on mainlines for 125-mph capable battery trains.

To start things off, these are my thoughts for main lines in my part of the world (i.e. the MML and the Cross Country route)

BEMU SPECIFICATION
It seems to me that the 10-car Class 897 trimode, or similar, as ordered by LNER would be a good candidate for modelling the OHLE needed for batterifcation of mainlines. The 897 has a claimed range of 120 miles on batteries, after which the diesel engines would need to be run. We don't have the full 897 specification, so I'm going to assume the following:

• Maximum range on battery alone: 120 miles
• Usual operating range (80% to 20% charge): 60% x 120 = 72 miles
• Time to recharge (20% to 80%): 18 minutes (This is equivalent to a "2C" recharge rate)
• Time to recharge (80% to 90%): 6 minutes (1C rate)
• Time to recharge (90% to 100%): 12 minutes (0.5 C)
• Maximum power draw: 6MW (with supplementary pantographs for use only when stationary or moving slowly)

INITIAL OVERHEAD ELECTRIFICATION
To start off, I'm assuming that we wire all sections of routes which carry 7 or more trains per hour in each direction. According to this thread, that means Nottingham-Trent East (7tph); Dore-Sheffield (8tph), Sheffield-Meadowhall (11tph). Also Bristol TM to Parkway (~7tph)

Trivia: What sections of unelectrified line carry the most trains per hour?

Considering lines that are not yet electrified, what sections carry the most trains per hour (tph)? At this stage I'm only considering timetabled passenger services, and not including freight or ECS/NR movements. Or station throats. I can think of the following: Dore - Sheffield (8 tph, I...

www.railforums.co.uk

Other wiring could be added as needed, but these would not be a pre-requisite for running 897s with their diesel capability.

NOTTINGHAM-LONDON
With several hours under the wires, all Down trains will pass Wigston with 100% charge. 25 miles to Trent East will take that down to ~80%, followed by at least 18 minutes to recharge (6 transit; 6 dwell; 6 transit), taking them back up to 100% again by Trent East on the way back south.

SHEFFIELD-LONDON
Wigston-Dore is 65 miles, easily within the operating range of the 897. Dore-Sheffield is ~5 minutes each way and most London trains have at least 15 minutes dwell or more at Sheffield. State of charge: Wigston 100%, Dore 45%, aSHF 62%, dSHF 90%, Dore 97%, Wigston 42%. Easily done without diesel.

EDINBURGH-PLYMOUTH
Southbound Cross Country services will be 100% charged at Moorthorpe, and will still have 87% charge by Meadowhall, restored to ~97% in the 12 minutes from Meadowhall to Dore. 76 miles to Birmingham (36%). 6 mins dwell at New St (56%) plus 12 mins to Kings Norton (85%); 75 miles to Westerleigh Jn (22%). 20 minutes under the wires will get back over 80% leaving Temple Meads, which is enough to get to Tiverton before having to start the diesels. Then run on Diesel and Battery to Plymouth and back to Temple Meads.

CARDIFF-NOTTINGHAM
Trains will leave Severn Tunnel Junction with 100% charge. 73 miles to Bromsgrove (39%); 6 mins to Longbridge (59%); 2 miles to Kings Norton (57%); 12 min to New St (77%), plus 6 min dwell (82%); 50 miles to Trent East (40%). At least 18 mins to recharge at Nottingham (~90+) before passing Trent East heading back south

CONCLUSION
So it seems to me that with trimodes similar to Class 897 you can batterify the entire MML mainline and Cross Country route as far as Tiverton with just 10 route miles of OHLE at Meadowhall-Dore, 7 miles Trent-Nottingham and 5 miles between Bristol Parkway and Temple Meads.

The key characteristic for the BEMU is a fast recharge rate under the wires. This dramatically shortens the length of wiring needed at each island of electrification.

Is this analysis realistic? What are your thoughts?
If you don't think it would work, what size battery would be needed?

A couple of major misconceptions:
[0. You would give any CAF engineers reading that a heart attack!!! and engineers elsewhere would spill their popcorn or coffee]
1. No one (UK or globally) is proposing BEMU operation above 100mph on battery as the energy usage is too high. DfT, NR, RSSB, operators and OEMs are all aligned on this.
2. 897 operation on diesel or battery is targeted on secondary routes where the maximum lines speeds are in the 65-75mph bracket with 70mph max specified off OHLE. E.g. Doncaster - Hull (~70miles return trip off wires and will use some diesel in reality and not just battery on this route this should tell you something about battery range! or rather lack thereof), Leeds - Harrogate (~34miles return trip off wires, battery only) and Newark - Lincoln ~33miles return trip off wires (battery only); and potentially but unlikely according to "Plan A" Northallerton - Middlesborough (still 801s in the future) with ~42.5 return trip miles off wires. The real range will be much lower than you state and under much friendlier conditions.
3. In practice 897 mileage ranges are based on operating at those lower speeds with lower energy requirements, not mileage at higher speeds. The CAF 397 data says resistance (davis eqn on the level) at 70mph is just 39% of 125mph, but as always there would be greater hotel / aux. usage per mile at lower speeds.
4. 897s will have very tame gentle battery usage cycles with at least 4hours (slow) charging between battery usage off wires. (What you are proposing isn't kind to batteries)
5. Aux /hotel usage during long layovers in the off wires termini will also nee to be accounted for
6. The only source for the 120mile range is "The Anonymous Widower" asking google ai in April 2026 and then writing it in his blog. His conclusion was that most of the google info was from a Railway Gazette /Rail business UK article in early 2026, however there is no mention of 120mile range in the article when I checked, hence 120miles range may be just an AI hallucination.

As @Zomboid correctly states a huge amount depends on battery chemistry and duty cycles
7. Rail usage clocks up charge and discharge cycles quickly hence you need to be really kind to batteries to get any decent working life out of them - operating any current batteries technology at those C rates will lead to a very short life. The most expensive LTO cells which are best optimises for cell life cycle and high C rates are only good to 1.7C max other chemistry cells shouldn't be taken much above 1C (for 20% --> 80% and even lower elsewhere) if you want to maximise cell life. (As cross sector comparison the new Tesla taxi is targeted to spend ~ half its time every day charging at moderate rates to be kind to the batteries)
8. Much more mileage will be need to be electrified than some people want to believe. Siemens advertorial in Modern Railway late last year was pretty much on the money roughly stating: "you need the right ~30% electrified as a minimum" and that is for regional rather than IC use.

897s and the new batch of Hitachi 8xx for First Lumo/Hull aren't a good benchmark for intercity modelling as they all target lower speeds and shorter of wires distances - they are targeted at the easy low hanging fruit.

[As someone with very good knowledge of Bombardier/Alstom, CAF, Hitachi, Siemens and Stadler rolling stock]

 

[GRALISTAIR]

1. No one (UK or globally) is proposing BEMU operation above 100mph on battery as the energy usage is too high. DfT, NR, RSSB, operators and OEMs are all aligned on this.

Power consumed is directly proportional to velocity cubed iirc? (Please trust me, I do know the difference between energy and power) - also related to resistance/friction/drag etc.

 

[hwl]

Power consumed is directly proportional to velocity cubed iirc? (Please trust me, I do know the difference between energy and power) - also related to resistance/friction/drag etc.

Yes. Force is proportional to v^2 (+ v + C + Potential Energy from elevation change) and Power is v^3 ([Force egn] *v), Energy is more complex on passenger trains as Hotel and aux loads /mile go up as you travel more slowly - the battery is best though of a bucket with a hole in it to represent hotel and aux loads, hence on battery there is an optimum compromise speed band to be ascertained between compromising for lower total power for hotel and aux if you go faster and lower total power due to dynamic resistances if you go slower (if you go too slowly traction system efficiency also drops which is another reason for not going to slowly!) Goldie Locks had lots of very relevant transferrable lessons on porridge to say on this!

 

[brad465]

On the XC network most of the 100+mph running is already electrified; Birmingham-Derby is the main exception, but this would be prime for electrifying anyway.

 

[Nottingham59]

Power consumed is directly proportional to velocity cubed iirc? (Please trust me, I do know the difference between energy and power) - also related to resistance/friction/drag etc.

Yes, it is - at high speeds where air resistance is the principal load. (But since the train gets there faster, energy consumed per mile is proportional to square of the speed.)

The best benchmark I know for the power required for high speed running is the Euston-Glasgow record attempt back in 2021, where a 9-car Pendolino used less that 8MWh to travel 400 miles in just under 4 hours. So an average consumption of less than 2MW.

https://www.modernrailways.com/article/lessons-record-run

ENERGY
Net energy used for the journey was 7,863kWh after regenerative braking returned 13% to the supply. This gave an average consumption of 12.2kWh/km,

 

[GRALISTAIR]

On the XC network most of the 100+mph running is already electrified; Birmingham-Derby is the main exception, but this would be prime for electrifying anyway.

You would think so -I am an electrification enthusiast. However, I will never on this forum use the dreaded phrase ---- en oh br ayn err !

 

[hwl]

On the XC network most of the 100+mph running is already electrified; Birmingham-Derby is the main exception, but this would be prime for electrifying anyway.

You would think so -I am an electrification enthusiast.

And some of the more problematic bridges at the Birmingham end are being dealt with by HS2.

== Doublepost prevention - post automatically merged: Yesterday at 14:03 ==

Yes, it is - at high speeds where air resistance is the principal load. (But since the train gets there faster, energy consumed per mile is proportional to square of the speed.)

The problems is the force available for acceleration decreases with speed hence increasing timing taken to accelerate to higher speeds means the energy consumed is greater than squared / unit distance when you look at real routes with non-steady state assumptions. On most GB routes /services substantial proportions of the overall distance are spent accelerating /decelerating rather than cruising at speed. Simplification is not kind to battery modelling unfortunately.

 

[GraphitProjekt]

Another consideration is existing power supply.

Southbound Cross Country services will be 100% charged at Moorthorpe,

Although the ECML recently underwent a PSU, it was unfinished north of Newcastle and is basically only capable to reasonable supply the last big December timetable change. The infrastructure on the ECML as it currently is would not be able to support XC trains even operating as 125mph EMUs let alone further draw for charging even if it's spread over the distance between Leeds and Edinburgh.

73 miles to Bromsgrove (39%); 6 mins to Longbridge (59%); 2 miles to Kings Norton (57%); 12 min to New St (77%), plus 6 min dwell (82%);

Similarly I highly doubt Winson Green and Galton Jnc feeders can handle any of that. Also depending on a dwell at New Street seems perhaps unwise when the timetable around there is so fragile, I know it's only a few percentage points but I wouldn't include it in a sketch like this. The power supply in the West Midlands in general is already a pinch point without any further electric traction than we have right now. Some intention exists from NR to address this put I haven't heard of even a plan yet.

With several hours under the wires, all Down trains will pass Wigston with 100% charge....
Trains will leave Severn Tunnel Junction with 100% charge.

It is indeed more likely that MML and GWML have electrical headroom available - but I still don't think it's at all certain that those lines can simply deliver the required power with no interventions. The majority of the MML is still classically fed although some passive provisions were made for retrofitting AT and all the feeder stations can provide that. It's also worth considering what the implications are of an N-2 situation at St Brides (Newport), or any of the 3 MML feeders (Braybrooke, Long Meadow Farm, Borehamwood)

with just 10 route miles of OHLE at Meadowhall-Dore, 7 miles Trent-Nottingham and 5 miles between Briston Parkway and Temple Meads.

But I raise this all because, no - you already need a lot more infrastructure investment then just that even for the favourable conditions this analysis relies on. PSU projects are really quite expensive although luckily somewhat straightforward.

2. 897 operation on diesel or battery is targeted on secondary routes where the maximum lines speeds are in the 65-75mph bracket with 70mph max specified off OHLE. E.g. Doncaster - Hull (~70miles return trip off wires and will use some diesel in reality and not just battery on this route this should tell you something about battery range! or rather lack thereof), Leeds - Harrogate (~34miles return trip off wires, battery only) and Newark - Lincoln ~33miles return trip off wires (battery only); and potentially but unlikely according to "Plan A" Northallerton - Middlesborough (still 801s in the future) with ~42.5 return trip miles off wires. The real range will be much lower than you state and under much friendlier conditions.

So perhaps on a good day, you could get from Doncaster to Hull and back off wires, but because realistically you can't depend on that, the whole fleet has to lug around all the diesel equipment as well even though the batteries are dependable for Harrogate and Lincoln routes. I mean, does it not seem a little absurd that we're building trains with diesel engines, fuel tanks, generators, traction batteries, a 25kV transformer and at least a pantograph all because we somehow can't do what the majority of continental railways did in 20th century. We know that bi-modes cost substantially more than EMUs upfront and will have higher maintenance costs too, tri-modes even more so. If you have to carry around a whole diesel engine, I can't personally understand how it's justifiable to also have the batteries - all just for platform air pollution? I guess the maintenance will be a bit cheaper if the engines are hardly used, but the track will nonetheless feel the brunt.

 

[Crithylum]

Taking the pendolino example of 8MWh, with an energy density of 160 w/kg (roughly middle of the range for new LFP or sodium batteries), that would require 50 tonnes of batteries.

At a cost of £100/kwh (which is on the high end for these chemistries), that comes out to £800,000.

I’m not going to call either of these negligible, however, in the grand scheme of things it is only a few percent of the total cost. This is to cross most of country at 125mph.

Batteries of these chemistries can achieve roughly 5000-10000 cycles (maybe more for future sodium ion batteries), so assuming ~3 runs per day, that is approximately 5-10 years lifespan. I would be surprised if the lifecycle cost was even a fraction of the upfront and maintenance cost increase of a DMU.

 

[Zomboid]

The issue isn't really battery capacity, it's charging speed (I doubt you're getting 5000 cycles at maximum charge rate without seriously degrading the end of life capacity) and conservation of energy.

Batteries aren't magic, they're an energy store. If it takes 8MWh to get from Glasgow to London in 4 hours, then if it's fully electrified that's an average of 2MW. If it's 50% wired (by journey time) then that system has to deliver 4MW whilst the train is drawing current. If it's 25% then the draw is 8MW. Multiply that by a few trains and you're taking a lot of power in a short time, the grid connections and the OLE have to be designed to handle that.

And that's before we consider the losses associated with energy conversion into and out of storage. Assuming 90% round trip efficiency (which is quite ambitious as I understand it), that 8MWh is now 8.8MWh. If the grid sites are battery backed to ensure the load is manageable by the grid, we're potentially putting that energy in and out of a battery twice, so 8MWh consumed could now be up to 9.8MWh.

Personally I think batteries have a big part to play in the future of traction power, but they are not the silver bullet that will solve every problem, especially when it comes to higher speeds and long distance operation.

 

[GRALISTAIR]

Another consideration is existing power supply.

Although the ECML recently underwent a PSU, it was unfinished north of Newcastle and is basically only capable to reasonable supply the last big December timetable change. The infrastructure on the ECML as it currently is would not be able to support XC trains even operating as 125mph EMUs let alone further draw for charging even if it's spread over the distance between Leeds and Edinburgh.

At an absolute minimum it would need Marshall Meadows to be SFC / AT to be installed and commissioned.

 

[2001olib]

Where does freight come into the plan. Busy freight corridors should be just as high up on the priority list as the passenger corridors. IMO all of the UKs major ports should have an electrified rail connection. There should also be another E-W electrified route, Peterborough-Nuneaton and onto Birmingham and Bescot springs to mind. The more freight becomes electric hauled with proper modern electric haulage timings the better, not from an eco stand point but from a pathing and capacity standpoint.

Other than people doing theoretical calculations on the internet batteries currently dont fulfil any of this.

 

[GRALISTAIR]

Pathing and capacity are certainly ( or should be) big considerations.

 

[gallafent]

On the XC network most of the 100+mph running is already electrified; Birmingham-Derby is the main exception, but this would be prime for electrifying anyway.

… and Didcot-Coventry/Birmingham (both routes north of Leamington!) … lots of XC traffic there too … and mostly 90mph (sigh, the days of the tilting 221s in the Cherwell Valley are now far behind us …).

From RailMiles:

Birmingham New Street — Derby: 40mi 36ch
Didcot East Junction - Coventry: 62mi 63ch
Didcot East Junction - Birmingham New Street (the "direct" route): 77mi 14ch

… and, yes, these routes are prime for electrifying anyway (especially chipping in from both ends, if we were still thinking about some battery operation). New Street to Leamington and Didcot to Oxford would certainly be a good start, and the former would be rather useful for Chiltern line services too.

 

[hwl]

Another consideration is existing power supply.

Another massive issue (of many

Although the ECML recently underwent a PSU, it was unfinished north of Newcastle and is basically only capable to reasonable supply the last big December timetable change. The infrastructure on the ECML as it currently is would not be able to support XC trains even operating as 125mph EMUs let alone further draw for charging even if it's spread over the distance between Leeds and Edinburgh.

Similarly I highly doubt Winson Green and Galton Jnc feeders can handle any of that. Also depending on a dwell at New Street seems perhaps unwise when the timetable around there is so fragile, I know it's only a few percentage points but I wouldn't include it in a sketch like this. The power supply in the West Midlands in general is already a pinch point without any further electric traction than we have right now. Some intention exists from NR to address this put I haven't heard of even a plan yet.

It is indeed more likely that MML and GWML have electrical headroom available - but I still don't think it's at all certain that those lines can simply deliver the required power with no interventions. The majority of the MML is still classically fed although some passive provisions were made for retrofitting AT and all the feeder stations can provide that. It's also worth considering what the implications are of an N-2 situation at St Brides (Newport), or any of the 3 MML feeders (Braybrooke, Long Meadow Farm, Borehamwood)

The power supply in the Reading area is pretty maxed out (no Bramley feeder). The reality is than large section of many exiting routes have constrained supply capacity

But I raise this all because, no - you already need a lot more infrastructure investment then just that even for the favourable conditions this analysis relies on. PSU projects are really quite expensive although luckily somewhat straightforward.

Agreed

So perhaps on a good day, you could get from Doncaster to Hull and back off wires, but because realistically you can't depend on that, the whole fleet has to lug around all the diesel equipment as well even though the batteries are dependable for Harrogate and Lincoln routes.

Which tells you that the "Google AI" search result in April 2026 of 120miles is probably a hallucination if you don't/can't trust it for just 60% of that.

I mean, does it not seem a little absurd that we're building trains with diesel engines, fuel tanks, generators, traction batteries, a 25kV transformer and at least a pantograph all because we somehow can't do what the majority of continental railways did in 20th century. We know that bi-modes cost substantially more than EMUs upfront and will have higher maintenance costs too, tri-modes even more so. If you have to carry around a whole diesel engine, I can't personally understand how it's justifiable to also have the batteries - all just for platform air pollution?

Have you ever been to New Street, Inverness or Sunderland? The BCR is poor as you would expect. One interesting point is that extended idling isn't not good for diesel engines (oil dilution) hence reducing the proportion of idling is beneficial overall.

I guess the maintenance will be a bit cheaper if the engines are hardly used, but the track will nonetheless feel the brunt.

 

[Zomboid]

Which tells you that the "Google AI" search result in April 2026 of 120miles is probably a hallucination if you don't/can't trust it for just 60% of that.

As with cars, any stated range is based on a whole bunch of assumptions (and in the case of cars, WLTP is not representative of how most people use their cars, EV or ICE).

So even if there is a non-hallucinatory figure of 120 miles out there for the 897, without understanding the assumptions behind that it doesn't help anyone. It might well be that they'll do more than 120 miles downhill or on level track at 35mph in 20*C weather with a tailwind. Or it might be more onerous, but I doubt you'll be doing 120 miles on a charge if the 125mph section between Birmingham and Derby is part of it.

 

[hwl]

Taking the pendolino example of 8MWh, with an energy density of 160 w/kg (roughly middle of the range for new LFP or sodium batteries), that would require 50 tonnes of batteries.

At a cost of £100/kwh (which is on the high end for these chemistries), that comes out to £800,000.

I’m not going to call either of these negligible, however, in the grand scheme of things it is only a few percent of the total cost. This is to cross most of country at 125mph.

Batteries of these chemistries can achieve roughly 5000-10000 cycles (maybe more for future sodium ion batteries), so assuming ~3 runs per day, that is approximately 5-10 years lifespan. I would be surprised if the lifecycle cost was even a fraction of the upfront and maintenance cost increase of a DMU.

The lack of overall rolling stock strategy is a big problem as the focus tend to be on direct replacement of old stock or short branch line because they should be easy (but have no easy sufficient charging opportunity in reality rather than biting off easier to BEMU routes and cascading non life expired rolling stock.

One good reason to hold off to 2030+ for delivery is that sensible high cycle life low flamability battery choices will become available and pricing will become much better too. Quick delivery is a case of more haste less speed.

== Doublepost prevention - post automatically merged: Today at 10:59 ==

The issue isn't really battery capacity, it's charging speed (I doubt you're getting 5000 cycles at maximum charge rate without seriously degrading the end of life capacity) and conservation of energy.

Batteries aren't magic, they're an energy store. If it takes 8MWh to get from Glasgow to London in 4 hours, then if it's fully electrified that's an average of 2MW. If it's 50% wired (by journey time) then that system has to deliver 4MW whilst the train is drawing current. If it's 25% then the draw is 8MW. Multiply that by a few trains and you're taking a lot of power in a short time, the grid connections and the OLE have to be designed to handle that.

And that's before we consider the losses associated with energy conversion into and out of storage. Assuming 90% round trip efficiency (which is quite ambitious as I understand it), that 8MWh is now 8.8MWh. If the grid sites are battery backed to ensure the load is manageable by the grid, we're potentially putting that energy in and out of a battery twice, so 8MWh consumed could now be up to 9.8MWh.

The DC-DC converter losses in the BMS increase at low power draw levels of 90 is very optimistic. The other thing to bear in mind with car comparisons is that passenger trains have much high aux and hotel loads per vehicle EMU of 32-40kW so a low end expectation for an electic only "897" would be 320-400kW. BEMU will be higher because of battery heating /cooling requirement, BMS etc. and Bi-modes on electric with engine warm up or cool down are considerably higher still (+15 to 70 kW per engine).

Personally I think batteries have a big part to play in the future of traction power, but they are not the silver bullet that will solve every problem, especially when it comes to higher speeds and long distance operation.

Completely agree. Unfortunately no one wants to focus on the more deliverable "boring" routes first (south of the border anyway)!

 

[zwk500]

Personally I think batteries have a big part to play in the future of traction power, but they are not the silver bullet that will solve every problem, especially when it comes to higher speeds and long distance operation.

100% Agree.

Where does freight come into the plan. Busy freight corridors should be just as high up on the priority list as the passenger corridors. IMO all of the UKs major ports should have an electrified rail connection.

Freight is going to need battery locos of some description for moving about the terminals, even if the A/D roads are electrified.

There should also be another E-W electrified route, Peterborough-Nuneaton and onto Birmingham and Bescot springs to mind. The more freight becomes electric hauled with proper modern electric haulage timings the better, not from an eco stand point but from a pathing and capacity standpoint.

Very much agree that the core electrification strategy should seek to provide a strategic freight network as well.

Other than people doing theoretical calculations on the internet batteries currently dont fulfil any of this.

Traction batteries will become part of the equation for freight soon enough. The Class 93s already have them, and I'd be surprised if the next big fleet were ordered with any diesel capability.

 

[GraphitProjekt]

Another massive issue (of many

The power supply in the Reading area is pretty maxed out (no Bramley feeder). The reality is than large section of many exiting routes have constrained supply capacity

I'm quite surprised that GWML is already maxing out electrical capacity since it's fully AT. Unlike MML which has the feeding distances (and grid connections) of an AT systems but without any of the required tracskide equipment ("saving money") apart for Braybrook - Wigston for some reason, and Kentish Town-Borehamwood which was a "Thameslink 2000" initiative.

I think there is actually a Bramley feeder these days. But the problem is that (in lieu of building the requisite electrification), the cable from Bramley to Reading was built with limited electrical capacity ("saving money"). I think it only comes out to play in the specific scenario that the railway is closed somewhere between Reading and Didcot, therefore Bramley can feed west of Maidenhead but probably only for Elizabeth line and Newbury services.

Which tells you that the "Google AI" search result in April 2026 of 120miles is probably a hallucination if you don't/can't trust it for just 60% of that.

The Anonymous Widower seems to have just made the number up, and for whatever reason his website seems reputable enough for google's ai.

But more to the point, it does suggest that even if we just build a small amount of electrification on the line to Hull, we could do away with the diesel engines and fuel tanks entirely on the fleet, which is surely a much better idea. (And then one remembers that TRUP originally proposed to electrify Micklefield/Temple Hirst to Hull.)

Considering the batteries are not dependable for this route, this suggests N-2 resilience would be required for the OLE which somewhat limits viability of just putting up a small amount of wires within the Hull area.

Instead, how many miles of OLE from Temple Hirst towards Hull would be needed to rid the diesel engines from the 897 fleet?

 

[zwk500]

Instead, how many miles of OLE from Temple Hirst towards Hull would be needed to rid the diesel engines from the 897 fleet?

Given the presence of the Selby swing bridge, one option would be to electrify Hull to Gilberdyke (16 miles), leaving about 20 miles off the juice through Selby. Such a project would allow a fairly natural roll-on to Goole and Doncaster, or up the coast to Beverley or Bridlington, and then leaves a nicely experienced team with local knowledge to wire Temple Hirst - Selby - Hambleton - Micklefield, with or without Church Fenton - Gascoigne Wood and some of the other more freight-dominated short links in the area.

 

[Zomboid]

Isn't there also a swing bridge at Goole? Whilst wiring moving bridges is a problem which has been solved across the world, that kind of thing does make sense as a battery use case if the only traffic is passenger. On it's own it'd only need a relatively small battery too.

 

[hwl]

I'm quite surprised that GWML is already maxing out electrical capacity since it's fully AT. Unlike MML which has the feeding distances (and grid connections) of an AT systems but without any of the required tracskide equipment ("saving money") apart for Braybrook - Wigston for some reason, and Kentish Town-Borehamwood which was a "Thameslink 2000" initiative.

I think there is actually a Bramley feeder these days. But the problem is that (in lieu of building the requisite electrification), the cable from Bramley to Reading was built with limited electrical capacity ("saving money"). I think it only comes out to play in the specific scenario that the railway is closed somewhere between Reading and Didcot, therefore Bramley can feed west of Maidenhead but probably only for Elizabeth line and Newbury services.

The Anonymous Widower seems to have just made the number up, and for whatever reason his website seems reputable enough for google's ai.

He stated he asked google ai because that was the only way he could find anything - which had ended up self-reinforcing the 120 number.

But more to the point, it does suggest that even if we just build a small amount of electrification on the line to Hull, we could do away with the diesel engines and fuel tanks entirely on the fleet, which is surely a much better idea. (And then one remembers that TRUP originally proposed to electrify Micklefield/Temple Hirst to Hull.)

Considering the batteries are not dependable for this route, this suggests N-2 resilience would be required for the OLE which somewhat limits viability of just putting up a small amount of wires within the Hull area.

Instead, how many miles of OLE from Temple Hirst towards Hull would be needed to rid the diesel engines from the 897 fleet?

Much better go to west from Hull to Gilberdyke as @zwk500 says and the OHLE can take the hotel and aux loads during the long layovers in Hull and charge at low-ish C rate that is kind to batteries and avoid the Selby headache.

== Doublepost prevention - post automatically merged: Today at 12:45 ==

Isn't there also a swing bridge at Goole? Whilst wiring moving bridges is a problem which has been solved across the world, that kind of thing does make sense as a battery use case if the only traffic is passenger. On it's own it'd only need a relatively small battery too.

yes hence zwk500 and I suggesting doing the far (Hull) side of Selby rather than the short ECML - Selby section. (This also make the TP Hull services and some Northern Service easy to BEMU...)

 

[zwk500]

Isn't there also a swing bridge at Goole? Whilst wiring moving bridges is a problem which has been solved across the world, that kind of thing does make sense as a battery use case if the only traffic is passenger. On it's own it'd only need a relatively small battery too.

Good point - yes, that would be a natural end point for electrification from Hull. It leaves just under 20 miles from Goole to Doncaster on Batteries (a bit less if you were to extend the wires 'around the corner' from Donny to help with acceleration away from the station or trains held at signals waiting for a platform)

 

[hwl]

Good point - yes, that would be a natural end point for electrification from Hull. It leaves just under 20 miles from Goole to Doncaster on Batteries (a bit less if you were to extend the wires 'around the corner' from Donny to help with acceleration away from the station or trains held at signals waiting for a platform)

It depends on which route is used for Doncaster to Hull, via Goole or via Selby both routes are used by different operators, East of Gilberdyke covers both.
Having the end point just east of the swing bridge would only leave a 6 mile gap in the via Selby route (NE of Templehirst Jn)

 

[zwk500]

It depends on which route is used for Doncaster to Hull, via Goole or via Selby both routes are used by different operators, East of Gilberdyke covers both.
Having the end point just east of the swing bridge would only leave a 6 mile gap in the via Selby route (NE of Templehirst Jn)

Yes, I would wire from Hull out as far as makes sense, which would likely be to the area of each moveable bridge but may be earlier if there's an awkward Level Crossing or something.

 

[Zomboid]

It depends on which route is used for Doncaster to Hull, via Goole or via Selby both routes are used by different operators, East of Gilberdyke covers both.
Having the end point just east of the swing bridge would only leave a 6 mile gap in the via Selby route (NE of Templehirst Jn)

ECML to Selby (from Temple Hirst and Hambleton Jn) only has a couple of bridges in Selby itself, so ought to be relatively straightforward to fit wires to. I suspect the station footbridge would be the hardest to resolve, but that at least is entirely railway so wouldn't require a long road closure.

 

[Nottingham59]

Taking the pendolino example of 8MWh, with an energy density of 160 w/kg (roughly middle of the range for new LFP or sodium batteries), that would require 50 tonnes of batteries.

At a cost of £100/kwh (which is on the high end for these chemistries), that comes out to £800,000.

I’m not going to call either of these negligible, however, in the grand scheme of things it is only a few percent of the total cost. This is to cross most of country at 125mph.

Batteries of these chemistries can achieve roughly 5000-10000 cycles (maybe more for future sodium ion batteries), so assuming ~3 runs per day, that is approximately 5-10 years lifespan. I would be surprised if the lifecycle cost was even a fraction of the upfront and maintenance cost increase of a DMU.

I think that illustrates the paradox that I'm trying to explore in this thread. Batteries are now cheap and getting cheaper as the technology develops, giving ever increasing capacity and faster charging times. But the rail industry seems to regard battery at best as a temporary aberation until "proper" electrification can be installed. I think that's a mistake.

No one (UK or globally) is proposing BEMU operation above 100mph on battery as the energy usage is too high.

It's only too high for sustained running over long distances. If the battery is big enough, and the distance short enough, surely the total energy used would be manageable? And remember there is not that much high speed running on the north MML. Trains take 66 mins (1.1h) to cover the 70 miles from Wigston to Sheffield; that's a average speed of 63mph.

897 operation on diesel or battery is targeted on secondary routes where the maximum lines speeds are in the 65-75mph bracket with 70mph max specified off OHLE. E.g. Doncaster - Hull (~70miles return trip off wires and will use some diesel in reality and not just battery on this route this should tell you something about battery range! or rather lack thereof),

Thank you. I hadn't realised that the battery on the 897 was so feeble. We need a bigger battery to model batterification properly, so I'd like to propose a different reference train, a 9-car IET with the following:

• 2 x 600kW diesel gensets
• 1 x 6MVA AC transformer
• Up to six 600 kWh battery packs (depending on route). 6 x 600kWh = 3.6 MWh: 80%-20% = 2.16 MWh.
• Recharge rate of 1.5C (so taking 24 minutes to recharge from 20% to 80% SoC)
• Three pantographs to allow charging at 6MW when stationary.

The Turntide battery used in the 802207 trial had a capacity of 575kWh, so this specification should be achievable with current technologies, even if LTO batteries are physically bigger than LFP. I'm going to call this configuration the 80x-6T

Given that London-Glasgow non-stop averages 2MW, I would expect average consumption from Wigston to Sheffield to be much less that 1.0 MW. So this 80x-6T train has enough capacity to get from Wigston to Sheffield and back again with no need to charge at all.

Similarly Moorthorpe to Derby is less than an hour. At 1MW average consumption that's less than 1MWh. From Derby to Birmingham takes 45 mins. At 2MW average, that's only another 1.5MWh. Achievable if the train is 100% charged at Moorthorpe, after several hours under the wires.

At an absolute minimum it would need Marshall Meadows to be SFC / AT to be installed and commissioned.

That would be good, but in a battery world upgrading Marshall Meadows is (surprisingly) not urgent. A BEMU could be programmed to switch to battery in the vicinity of Marshall Meadows, and only take recharge current again later in its journey at those places and times where supply is not constrained.

Even better, the train could be programmed to normally use AC, but switch to battery when the line voltage drops below a certain threshold, say 23kV.

The ability of Battery trains to time-shift their power take means that a BEMU rail network needs very much less total grid supply capacity than a rail network running pure EMUs. That is a huge saving in capital costs, especially nowadays, when grid supply points are so constrained.

If it's 25% then the draw is 8MW. Multiply that by a few trains and you're taking a lot of power in a short time, the grid connections and the OLE have to be designed to handle that.

Yes, it does. But it is much more efficient to place that grid supply at heavily used nodes in the network, where it can be used to recharge large numbers of battery trains continuously over the course of an hour. Instead of places like Tweedbank which will only ever see two trains per hour max.

Note that the number of electrification islands you need varies as the square of the range of the BEMUs being used. If you double the operating range of all BEMUs then you only need one quarter as many recharging nodes. (In the same way as a grid of 25x25km squares has four times as many nodes as a grid with 50km squares. )

 

[HSTEd]

So perhaps on a good day, you could get from Doncaster to Hull and back off wires, but because realistically you can't depend on that, the whole fleet has to lug around all the diesel equipment as well even though the batteries are dependable for Harrogate and Lincoln routes. I mean, does it not seem a little absurd that we're building trains with diesel engines, fuel tanks, generators, traction batteries, a 25kV transformer and at least a pantograph all because we somehow can't do what the majority of continental railways did in 20th century. We know that bi-modes cost substantially more than EMUs upfront and will have higher maintenance costs too, tri-modes even more so. If you have to carry around a whole diesel engine, I can't personally understand how it's justifiable to also have the batteries - all just for platform air pollution? I guess the maintenance will be a bit cheaper if the engines are hardly used, but the track will nonetheless feel the brunt.

Unfortunately the situation we are in is that mass 25kV electrification has become too expensive to be workable.

At £2.5-4m per single track kilometre, we can put an awful lot of equipment on a train before it becomes cheaper to pay that.
Even if we have 25kV transformers, batteries and a diesel engine in a trainset, it will be a long time before we would be better off coughing up for the 25kV.

Most likely I coudl see a situation where the trains operate as battery electric multiple units 99% of the time, but if there is some feeder fault or other they turn on their diesel engines and run them continuously to either reduce load on the electrification system or slow battery discharge. 560-700kW running continuously will stretch a battery an awful lot.

A battery electric EMU only needs to draw its average power draw from the traction supply, but will draw it continuously. And (as noted above), it can stop drawing if the grid shows voltage drop stress. A Pendolino sized unit won't suddenly spike demand to 6MW, it will just plug along at 2MW or so - so we can probably tolerate lower headrooms.