Modelling performance and range of Hitachi 80x Trimode trains, and implications for future mainline electrification

[Nottingham59]

BATTERY TRAIN ENDURANCE
When thinking about electrifying the rest of the GB mainline network, it seems to me that one key constraint is going to be the performance and range of 125mph-capable battery units. I have three hard pieces of evidence to help model this; if anyone has any better data, then please share it:

1. The battery trial with 802207, where one of the three 700kW diesel generating units (GUs) was swapped for a 700kW/550kWh battery module supplied by Turntide. Performance was the same on battery and diesel, fuel consumption off the wires reduced by 30-50%. The range on battery alone was 43 miles. For details see: https://www.railforums.co.uk/thread...r-for-transpennine.249321/page-3#post-7023353
2. The original government specification for the Intercity Express Programme (IEP) was that a 9-car 234m IEP should be able do Kings Cross to Newcastle with seven stops in 162 minutes, using no more than 8120kWh. This equates to an average power consumption of 3.03MW on a 125mph railway. See: https://assets.publishing.service.gov.uk/media/5a79b82e40f0b63d72fc7f58/tts-redacted.pdf
3. First group have ordered 14 x 5-car battery bimodes and battery/diesel trimodes from Hitachi for £300M for their open access operation, but I haven't seen any performance specs. The Turntide contract was for £10M, but included development work too.

REFERENCE TRAIN
So I've chosen as a reference train a 9-car Hitachi 80x trimode ("80x-T") with:

• 2 x 4MW transformers, one in each driving car. Each end car to have two pantographs, so that the train can recharge its batteries at 8MW in a station without exceeding the 2MW limit for OHLE current draw when stationary. When moving, just one pantograph could handle 8MW.
• 2 x 700kW diesel generating units, for redundancy and extended range off the wires.
• 5 x 700kW / 550kWh battery packs, using existing Turntide technology, so total battery capacity of 2.75MWh.

Power output will be 3.5MW on battery alone and 7x700kW = 4.9MW peak, enough to keep to 80x or 810 timings on a mainline.

Given 3MW average energy consumption on a 125mph railway, I'm going to guess 50kW hotel power and auxiliary power per car (450kW) and therefore 2.5MW average traction power, which at 125 mph will be mostly wind resistance. Wind resistance varies as the cube of the speed, so this implies power consumption as follows:

• on a 125mph railway: 0.45MW Hotel/auxilliary + 2.58MW traction = 3.03MW average. Endurance on battery = 54 minutes
• on a 110mph railway: 0.45MW Hotel/auxilliary + 1.76MW traction = 2.11MW average. Endurance = 78 mins
• on a 90mph railway: 0.45MW hotel/auxilliary + 0.96MW traction = 1.41MW average. Endurance = 117 mins

MIDLAND MAINLINE
Applying these figures to MML electrification, I reckon the MML is a 125mph railway from the limit of wiring at Wigston to East Midland (EMD), and a 90mph railway from EMD to Nottingham. A 222 meridian covers this these segments in 19 and 8 minutes respectively, so I model 80x-T consumption as 0.96MWh + 0.19MWh = 1.15MWh. A 80x-T should get from Wigston to Nottingham and back again on battery alone, and have 0.45MWh of battery capacity remaining to supply 60 minutes of full hotel power sitting at Nottingham station.

On the route to Sheffield, I reckon that EMD to Clay Cross is a 110mph railway (2.11MW for 30 minutes = 1.05MWh), and Clay Cross to Sheffied is at best 90mph (0.96MW for 16mins = 0.25MWh). So energy consumption from Wigston to Sheffield station models at 0.96+1.05+0.25 = 2.26MWh, well within the battery capacity of the 80x-T.

Wiring just the platforms at Sheffield would allow the train to recharge in 2.26/8 hours = 17 minutes. That's less than the typical layover time at Sheffield. But if they needed to turn round faster than that, then you'd just run one or both diesel generating units. Alternatively, wiring the line from Sheffield to Dore (6 minutes away) would give 12 minutes to recharge on the move and need just 5 at the platform to recharge fully.

To recharge up to three 80x-Ts simultaneously at Sheffield I would wire platforms 5, 6, and 8 at Sheffield station, with rigid conductor bars and fed from a static frequency converter (SFC) rated at 25MVA continous power, fed from the Attercliffe 132kV substation next to the railway on the line out to Rotherham.

CONCLUSION
I reckon that all London services on the Midland Mainline could be converted to battery operation, using existing battery technology, simply by wiring the platforms at Sheffield station with thick conductor bars, and using 9-car 2.75MWh trimode trains, with diesel generating units to provide back up when needed but not used in normal operation.

What are your thoughts? Do you know of any better performance and energy consumption data I could use?

 

[edwin_m]

Wind resistance varies as the cube of the speed

The most common formula for calculating resistance is the Davis Equation, with coefficients for the zeroth, first and second powers of speed only. Air resistance is generally reckoned to vary with the square of speed.

Rail vehicle resistance - Wikipedia

en.wikipedia.org

A number of experimental measurements [3][4][5] of the train resistance have shown that this force can be expressed as a quadratic equation with respect to speed as shown below:

 

[Nottingham59]

The most common formula for calculating resistance is the Davis Equation, with coefficients for the zeroth, first and second powers of speed only. Air resistance is generally reckoned to vary with the square of speed.

Rail vehicle resistance - Wikipedia

en.wikipedia.org

Agreed. The force varies as the square of the speed, but power is force times speed, so the power needed varies as the cube.

So if you double the speed, you need four times the force but eight times the power. (Though you will get to your destination in half the time, so the overall energy consumed will only be four times).

EDIT:
But thanks for the Wikipedia link. I'd not found that article before. Pity that it doesn't give typical values for the coefficients, and all the references it cites are behing paywalls.

But I did find this article: https://www.modernrailways.com/article/lessons-record-run
A 9-car Pendolino on a record run used 7,863kWh (12.2kWh/km) between London and Glasgow.

This column’s status as the official publication of the militant wing of the Railway Electrification Front meant I received details of the ninecar Pendolino’s energy consumption. 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, which compares with the 14kWh based on measured readings between London and Manchester quoted in Roger Kemp’s traction metrics study for the RSSB. Informed sources confirm that in normal (pre-Covid) service a nine-car Pendolino was averaging around 14kWh/km, with regenerative braking savings of around 15%.

 

[edwin_m]

Agreed. The force varies as the square of the speed, but power is force times speed, so the power needed varies as the cube.

So if you double the speed, you need four times the force but eight times the power. (Though you will get to your destination in half the time, so the overall energy consumed will only be four times).

OK, understood if expressed in terms of power.

I have a spreadsheet that models train performance over increments of time, so if I was to extend it to record energy consumption I would be adding the resistance force times distance travelled over the time increment plus an item for any acceleration or regenerative braking.

 

[Nottingham59]

A 9-car Pendolino on a record run used 7,863kWh (12.2kWh/km) between London and Glasgow.

At that rate, a 2.75MWh 9-car battery Pendolino would have a range of 2750/12.2 = 225km. Just enough to cover the 70 miles from Wigston to Sheffield and back.

 

[Brubulus]

Though if we are talking about the MML, shouldn't we think of putting batteries in 810s?

The Turntide batteries would not be viable for long-term rail use, they were Nissan Leaf batteries, which are notorious for poor thermal management and fast degradation. Modern LFP batteries would have to be used, but assuming a weight per battery pack of 6 tonnes and an energy density of 135wh per KG, that enables around 800kw.

From what I've heard, power consumption on a 5 car 801 at 100mph is around 2.2kwh per vehicle mile. Given it's around 70 miles from Wigston to Sheffield this is around 770kw in normal conditions. Assuming a 20% buffet for increased consumption, this means usage would be around 936kw. 60% of 2 batteries is around 960kw. This means that 2 batteries would be acceptable for 810s.

Pendolinos will probably need more power than equivalent units for to tilting, and yes, deeming the MML north of Wigston a 100mph railway might be giving optimistic numbers, so that's what the 20% buffer is for. Secondly 20-80% is a very conservative battery use. 80% is probably necessary for quick charging, but on the lower end, 15% is probably acceptable in poor weather conditions, which can degrade down to 15% in normal operation and 10% in poor weather for it to really become a concern.

When weather is predicted to be very poor, the battery could be charged to 90% by Wigston, which effectively unlocks another 5%(for each direction). Both of these together unlocks 10% degregation, which means that 5-7 years of intensive use is likely possible with such batteries until they are moved to less strenuous tasks.

On that, it should be set up so batteries are interchangeable across the 80x family so that each battery set can be used twice, to give a lifespan of 15 years or so.

 

[Nottingham59]

Though if we are talking about the MML, shouldn't we think of putting batteries in 810s?

You'd have thought so, but the 810s have a different underfloor layout, and Turntide's new battery packs have been developed for the 80x. So I'm just modelling a train that can use the current Turntide design ....

The Turntide batteries would not be viable for long-term rail use, they were Nissan Leaf batteries, which are notorious for poor thermal management and fast degradation. Modern LFP batteries would have to be used, but assuming a weight per battery pack of 6 tonnes and an energy density of 135wh per KG, that enables around 800kw.

I understand the Leaf batteries were just for the trial. The current Turntide designs are LFP batteries rated at (I believe) 700kW power and 550kWh capacity, or better.

From what I've heard, power consumption on a 5 car 801 at 100mph is around 2.2kwh per vehicle mile. Given it's around 70 miles from Wigston to Sheffield this is around 770kw in normal conditions. Assuming a 20% buffet for increased consumption, this means usage would be around 936kw. 60% of 2 batteries is around 960kw. This means that 2 batteries would be acceptable for 810s.

Looks about right. I'm actually quite shocked by what these numbers are telling me: any future electrification of the MML will be redundant except for the platforms at Sheffield (and just maybe short heavily used sections like Sheffield-Dore and Nottingham-Derby.)

Pendolinos will probably need more power than equivalent units for to tilting, and yes, deeming the MML north of Wigston a 100mph railway might be giving optimistic numbers, so that's what the 20% buffer is for. Secondly 20-80% is a very conservative battery use. 80% is probably necessary for quick charging, but on the lower end, 15% is probably acceptable in poor weather conditions, which can degrade down to 15% in normal operation and 10% in poor weather for it to really become a concern. When weather is predicted to be very poor, the battery could be charged to 90% by Wigston, which effectively unlocks another 5%(for each direction). Both of these together unlocks 10% degregation, which means that 5-7 years of intensive use is likely possible with such batteries until they are moved to less strenuous tasks.

That's why I am a fan of tri-modes. Size the battery so that it meets all needs in normal use, but have a diesel engine in reserve for that one occasion per year when it's needed. A 800kWh of battery may weigh 6t, but 800kWh of diesel fuel will only weigh 240kg at 30% generation efficiency.

On that, it should be set up so batteries are interchangeable across the 80x family so that each battery set can be used twice, to give a lifespan of 15 years or so.

Hadn't thought of it that way, but it's a great idea. Thanks for all the feedback

 

[Bald Rick]

A great speculation thread!

Are you forgetting regen braking into the batteries? The consumption figures you quoted are gross (I think).

Personally I think the diesel engines in your putative train are superfluous. It is extra weight, and more costly maintenance capability (kit and people), for minimal benefit. It would be much more simple just to put another battery pack in for a bit more range, and/or have a ‘limp mode’ to get a train back onto the wire at slower speed and with reduced hotel demand if the juice runs low.

I’m now looking forward to similar calculations for a twin Class 99 loco with no diesel engines but 80 tonnes of LFP or LiTo batteries across the pair.

 

[Trainbike46]

On the Trimode Vs pure BEMU issue: if you replaced the space/weight of the engines with more batteries, you'd have more than enough spare battery capacity than you'd ever need.

 

[Nicholas Lewis]

A great speculation thread!

Are you forgetting regen braking into the batteries? The consumption figures you quoted are gross (I think).

Personally I think the diesel engines in your putative train are superfluous. It is extra weight, and more costly maintenance capability (kit and people), for minimal benefit. It would be much more simple just to put another battery pack in for a bit more range, and/or have a ‘limp mode’ to get a train back onto the wire at slower speed and with reduced hotel demand if the juice runs low.

I’m now looking forward to similar calculations for a twin Class 99 loco with no diesel engines but 80 tonnes of LFP or LiTo batteries across the pair.

Regen not so valuable for long distance trains with infrequent stops though.

What Ive yet to see is what provision a trains needs to have to deal with operating incidents or will they be treated like any other EMU ie how much power needs to be held back in reserve thus total available for traction wont be full rating of battery.

 

[Brubulus]

Regen not so valuable for long distance trains with infrequent stops though.

What Ive yet to see is what provision a trains needs to have to deal with operating incidents or will they be treated like any other EMU ie how much power needs to be held back in reserve thus total available for traction wont be full rating of battery.

Hotel power requirements for modern trains are pretty low. Across a 5 car unit, I'd budget 2kw for lighting, 8kw for passenger needs, 10kw for heating/cooling and 5kw for other needs (train systems, toilets etc). 25kw per hour.

3 hours of emergency reserve is reasonable for almost all situations, so 80kw can be allocated. Weather can be predicted in advance and it's impact on range should be known from testing and much less severe than on a car, as battery heating systems and similar have less of an effect. Diversions can obviously be predicted in advance and this might be an issue, but no more so than in an EMU.

 

[Bald Rick]

Regen not so valuable for long distance trains with infrequent stops though.

Agreed, but it is still wort something. EMR north of Leicester do tend to stop relatively frequently.

 

[Nottingham59]

What Ive yet to see is what provision a trains needs to have to deal with operating incidents or will they be treated like any other EMU ie how much power needs to be held back in reserve thus total available for traction wont be full rating of battery.

That's why I'm a fan of tri-modes. A tank of diesel is a far more efficient as a battery that will only be needed once a year. And diesel engine maintenance depends on running hours. If they don't run, they don't need much maintenance.

Regen not so valuable for long distance trains with infrequent stops though.

But stil the Pendolino example above returned 13% of its energy back to the system by regeneration. Not as much as a suburban metro with stop-start running, though.

Hotel power requirements for modern trains are pretty low. Across a 5 car unit, I'd budget 2kw for lighting, 8kw for passenger needs, 10kw for heating/cooling and 5kw for other needs (train systems, toilets etc). 25kw per hour.

3 hours of emergency reserve is reasonable for almost all situations, so 80kw can be allocated. Weather can be predicted in advance and it's impact on range should be known from testing and much less severe than on a car, as battery heating systems and similar have less of an effect. Diversions can obviously be predicted in advance and this might be an issue, but no more so than in an EMU.

Thanks for those figures. I'll plug them into the model and see what comes out. Do you have any data for background energy consumption by the traction systems, i.e. not related to distance travelled?

 

[Bald Rick]

That's why I'm a fan of tri-modes. A tank of diesel is a far more efficient as a battery that will only be needed once a year. And diesel engine maintenance depends on running hours. If they don't run, they don't need much maintenance.

I disagree. For very infrequent use they are more trouble than they are worth.
unless absolutely necessary. Safe to assume that if battery technology had been sufficiently developed at the time, LNERs 801s would have a battery for emergency traction rather than a solitary engine.

But stil the Pendolino example above returned 13% of its energy back to the system by regeneration.

Indeed, and on the EMR IC services with a more choppy speed profile, lower average speeds, and more stops in many services, it will be higher than that.

 

[Nottingham59]

I disagree. For very infrequent use they are more trouble than they are worth.
unless absolutely necessary.

Well the ideal is a flexible platform where diesel GUs can be swapped for battery packs and vice versa. ....

Safe to assume that if battery technology had been sufficiently developed at the time, LNERs 801s would have a battery for emergency traction rather than a solitary engine.

Which rather begs the question why LNER 801s do not already have a programme in place today to progressively replace GUs with batteries. Maybe they do, but it hasn't been announced yet.

I can see a really good business case for all GWR 80x having one diesel unit swapped out for a battery. But I guess the DfT-mandated contract with Hitachi wouldn't cope with such flexibility.

 

[Trainbike46]

Which rather begs the question why LNER 801s do not already have a programme in place today to progressively replace GUs with batteries. Maybe they do, but it hasn't been announced yet.

I can see a really good business case for all GWR 80x having one diesel unit swapped out for a battery. But I guess the DfT-mandated contract with Hitachi wouldn't cope with such flexibility.

you give the likely answer to your first point in your second...

 

[HSTEd]

The likely reason battery packs have not yet been ordered for the entire 80x fleet is the same issue that causes many problems elsewhere on the railway.

Contractual paralysis.

 

[andrewgs]

8 MW for 2.75MWh pack
Leaf MY14-17 https://ev-database.org/uk/car/1019/Nissan-LEAF-24-kWh batteries had a max charge rate of 46 kW, which reduces with SOC. DC charging between 10% to 80% averages at 36 kW.
Scaling the Leaf's charging power of 36 kW for a 24 kWh pack to your proposed 2.75MWh gives a charge rate of 4.125 MW, roughly half your 8 MW assumption, so doubling your layover time. Note that newer cars have faster charging rates relative to their capacity.

> Wiring just the platforms at Sheffield would allow the train to recharge in 2.26/8 hours = 17 minutes


The duty cycle for a train doing multiple (3?) round trips, so 6 charge cycles per day with rapid charging is very different to an automotive duty cycle of 1 charge cycle a day (the UK average mileage gives 1 charge cycle/week for new EVs!).

Three years ago I worked on similar battery projects and remember 30-70% and 20-80% being considered, with allowances for emergency use below 20 or 30%. I would expect to size the batteries based on 20-80% on a winter day after the battery towards the end of its lifespan, after its has had several years of degradation.

My intuituion is that using space/weight for additional batteries is better than for a range extending diesel engine as you'll be able to run the batteries with smaller charge cycles, reducing degredation.

 

[edwin_m]

Well the ideal is a flexible platform where diesel GUs can be swapped for battery packs and vice versa. ....


Which rather begs the question why LNER 801s do not already have a programme in place today to progressively replace GUs with batteries. Maybe they do, but it hasn't been announced yet.

I can see a really good business case for all GWR 80x having one diesel unit swapped out for a battery. But I guess the DfT-mandated contract with Hitachi wouldn't cope with such flexibility.

Installing and maintaining a battery pack may be cheaper over its whole life than installing and maintaining a diesel engine and associated equipment. That doesn't mean it's cheaper than throwing away a diesel engine and associated equipment that's already bought and paid for.

 

[Narrowgauge725]

It has always seemed real strange to me that they chose Leaf batteries for the test, these are known not to be the best and are very early traction batteries with no thermal management. Battery technology is moving at a very great rate since Leaf batteries were designed. Other batteries are significantly less degradation and the future solid state batteries are likely to have very little degradation.

 

[HSTEd]

It has always seemed real strange to me that they chose Leaf batteries for the test, these are known not to be the best and are very early traction batteries with no thermal management. Battery technology is moving at a very great rate since Leaf batteries were designed. Other batteries are significantly less degradation and the future solid state batteries are likely to have very little degradation.

I believe the packs were considered to have the best safety record of those that were easily available for the trial.

 

[Nottingham59]

I believe the packs were considered to have the best safety record of those that were easily available for the trial.

I'm sure Turntide knew what they were doing. Even crappy old Leaf batteries gave 43 miles range off the wires, and demonstrated a 50% fuel saving. And that got them a £10M contract from Arriva Group for their 9-train Grand Central order, including covering some of their development costs.

Turntide Secures Hitachi Order to Power First UK Battery Trains

Turntide Secures Hitachi Order to Power First UK Battery Trains. This latest development strengthens their long-term relationship.

turntide.com

 

[Bald Rick]

It has always seemed real strange to me that they chose Leaf batteries for the test, these are known not to be the best and are very early traction batteries with no thermal management. Battery technology is moving at a very great rate since Leaf batteries were designed. Other batteries are significantly less degradation and the future solid state batteries are likely to have very little degradation.

The point being that if it worked with old Leaf batteries lashed together (which could be done quickly as a proof of concept), then a future rail spec design with batteries more suited to the charge/discharge cycle of rail applications would be rather better.

 

[eldomtom2]

Personally I think the diesel engines in your putative train are superfluous. It is extra weight, and more costly maintenance capability (kit and people), for minimal benefit. It would be much more simple just to put another battery pack in for a bit more range, and/or have a ‘limp mode’ to get a train back onto the wire at slower speed and with reduced hotel demand if the juice runs low.

Well, First and LNER still seem to consider diesel engines necessary.

 

[Bald Rick]

Well, First and LNER still seem to consider diesel engines necessary.

But they were specified over a decade ago, and not by First or LNER. And not for the train this thread is about.

 

[Zomboid]

I think this is looking at too narrow a problem. Batteries may well be suitable for the MML Intercity operation as described (though they won't be making significant changes to the 810s in a hurry), but broadly speaking that won't do much to help de-dieselise any other services.

The actual network solution will need to consider Cross Country through Derby & Sheffield, plus the local services in the Sheffield, Leicester and Nottingham areas, and I doubt that some conductor beam in Sheffield station is going to solve everything.

 

[Nottingham59]

I think this is looking at too narrow a problem. Batteries may well be suitable for the MML Intercity operation as described (though they won't be making significant changes to the 810s in a hurry), but broadly speaking that won't do much to help de-dieselise any other services.

Probably not. Station conductor bars only work for terminating services, with adequately long layovers. Even with 8MW recharge rates.

The actual network solution will need to consider Cross Country through Derby & Sheffield, plus the local services in the Sheffield, Leicester and Nottingham areas, and I doubt that some conductor beam in Sheffield station is going to solve everything.

That's for another thread. I was going to start it, but realised that I neeeded a handle on the potential range and performance of a 125-capable BEMU first. The many thoughtful and informative responses to this thread really have helped here.

(My best guess is currently that EMD to DBY and Dore to Meadowhall will be enough to electrify many services across the East Midlands. Adding wires where lines cross the ECML at Retford, Newark Castle and Grantham would cover most of the rest. But that just a guess at this stage. It all depends on the range of a BEMU to replace Class 170s.)

 

[HSTEd]

I think this is looking at too narrow a problem. Batteries may well be suitable for the MML Intercity operation as described (though they won't be making significant changes to the 810s in a hurry), but broadly speaking that won't do much to help de-dieselise any other services.

The actual network solution will need to consider Cross Country through Derby & Sheffield, plus the local services in the Sheffield, Leicester and Nottingham areas, and I doubt that some conductor beam in Sheffield station is going to solve everything.

I'd expect it'd come a lot closer than you might think.

 

[eldomtom2]

But they were specified over a decade ago, and not by First or LNER. And not for the train this thread is about.

I was referring to the 897s and Lumo's order. Apologies if that was unclear.

 

[Nottingham59]

A 9-car Pendolino on a record run used 7,863kWh (12.2kWh/km) between London and Glasgow.

That's an average consumption of 33.7kWh/min = 2.025MW.

Hotel power requirements for modern trains are pretty low. Across a 5 car unit, I'd budget 2kw for lighting, 8kw for passenger needs, 10kw for heating/cooling and 5kw for other needs (train systems, toilets etc). 25kw per hour.

So 5kW per car. Thanks.

Personally I think the diesel engines in your putative train are superfluous.

OK. So if you won't let me have diesel engines, let's try a 9-car all battery Hitachi IET, the 80x-B

• 2 x driving cars with 6MW transformer under the floor and three pantographs for use at stations with 25kV conductor bars sufficiently thick to deliver 6MW to each end car.
• 7 x Battery packs rated at 700kW / 550kWh. Each weighing the same as an 80x diesel GU.
• Hotel power = 5 x 9 = 45kW
• Maximum power output 4.9MW,
• Total battery capacity = 3.85MWh. At WCML record attempt consumption, that's 1.9h = 114 minutes endurance.

Assuming the MML north of Wigston and the Cross Country route both use power at the same 2MW average rate as a Pendolino to Glasgow:

• Wigston to Sheffield (67 mins) will use 59% of the battery. e.g. 90% to 31%. Ten minutes recharging at 12MW at Sheffield will put back 2MWh (52%). So the return journey will get back to Wigston with 24% charge left. If Sheffield grid supply fails, then the return journey will have to be in limp home mode (but wiring Derby station too would fix that).
  
  
• On the Cross Country route, Proof House Junction near New St to South Kirby Jn near Wakefield (89 mins) will consume 78% of the battery. e.g. 90% to 12%, without recharging at all. Five minutes dwell at Sheffield (or Derby if wired) would give a much bigger reserve. (Or you could wire Tamworth, fed from the WCML grid feeder point already there.)

In practice, I'd expect MML and XC consumption to be less than a Pendolino being driven to the limit, but that's the best I can do for now without more precise data.