Baffling Battery Blunder:
Expensive, experimental, slow battery locomotives have no place on already electrified rails
March 13, 2025
Against international best practices, the MTA is planning on procuring expensive, experimental battery-powered versions of Siemens Charger SC-42 locomotives [1].
Baffling and Irresponsible
The MTA’s vote on Wednesday to buy experimental battery locomotives for Penn Station Access [2] (PSA) is fiscally irresponsible, and threatens to hamstring our regional rail network’s ability to deliver frequent, fast, and reliable service for decades into the future. The MTA has inexplicably chosen to acquire battery-electric Siemens Charger locomotives, locomotives that cost an astounding $23.45 million each [3]. That would make these locomotives not only the first passenger battery locomotive in the world, but also the most expensive as far as we are aware—costing even more than an entire 8-car Metro-North M8 train [4]. And the reason for this expense? So they can run these locomotives in battery mode on tracks that are already electrified in and around Penn Station.
This is baffling and illogical. There are multiple superior, proven train options that can handle the different types of electrification on PSA. Improving connectivity between Midtown West, the Bronx, and Southwestern Connecticut is only possible with the right trains, but battery locomotives will condemn new Bronx riders to slow, unsatisfactory service, and worsen existing service on the New Haven Line. Furthermore, the rationale for acquiring new trains is inscrutable, given that the MTA already purchased M8s for PSA [5].
ETA urges the MTA to instead follow global best practices for modernized regional rail services and reverse course, switching to using the electric multiple units (EMUs) they already bought (the M8s). The MTA should follow plain common sense and simply run electric trains on already electrified tracks.
New Service along an Already Electrified Route
An Amtrak Northeast Regional pulled by a Siemens ACS-64 locomotive running under 12.5 kV 60 HZ overhead catenary, just north of Gate Interlocking on the Hell Gate Line. Image from Amtrak on their Penn Station Access Infrastructure Project page.
Penn Station Access will see Metro-North service run from New Rochelle station on the New Haven Line into Penn Station over the Hell Gate Bridge, following existing Amtrak tracks, with opening planned for 2027. The entire route is electrified, and electric Amtrak Acelas and Northeast Regionals run on this route every day—without the need for expensive, heavy batteries. Metro-North should have no need for batteries to run on already electrified tracks—it would be like charging your phone at home by plugging it into your car’s battery rather than into the wall.
That said, there are some complexities to the way the Penn Station Access route is electrified, and to understand them better, it helps to first understand the different electrification systems in place along the New Haven and Hell Gate Lines that it will follow, shown below:
Map of existing mainline electrification standards in the Tri-State region — Kara Fischer, ETA.
The Northeast has five different electrification standards. Three are with overhead catenary wires, used in New Jersey and on the Northeast Corridor. Two are with third rail, with the LIRR using over-running third rail shoes and Metro-North using under-running ones. The M7s and M8s can run under either kind of third rail, but can only be changed from one to another in the shop, not on the fly. The M8s can also run under catenary, but only if it is with 60 Hz frequency rather than 25 Hz. The reason for this has to do with on-board transformers: AC electrification requires the train to have transformers of size inversely proportional to the frequency, so that every 25 Hz train can also run under 60 Hz with little difficulty, but a 60 Hz train would need to be redesigned with larger transformers. When the M8 was first planned in the 2000s, it was only built with a large enough transformer for 60 Hz because of space and weight limitations. Many trains, including some owned by Amtrak, NJ Transit, and SEPTA, can easily operate on both, but to do so, they require onboard transformers that can convert both forms of electricity [6,7].
Or, more precisely, the M8 can run almost to New York, due to limitations in the Penn Station area:
A screenshot from OpenRailwayMap showing the various types of electrification (labeled) in Queens on the Northeast Corridor (NEC) and around Penn Station. The left shows Penn Station and West Side Yard. The right shows Sunnyside Yard, Harold Interlocking, Gate Interlocking, and the Hell Gate Line. For more information on the technical standards of electrification in this map, see [7].
Given this situation, Metro-North’s previous plan was straightforward and logical: it would serve Penn Station Access with the dual-voltage M8 trains it already operates by extending LIRR-type third rail from Harold to Gate Interlocking to bridge the gap between third rail and 60 Hz catenary. This solution is in fact already under construction [8]. This would enable M8s to run from the Hell Gate Line all the way into Penn Station and the West Side Yard, where they can use the third rail on any accessible track, just like LIRR trains.
While third rail is much more expensive than catenary to build and maintain [9], this is a sensible plan. EMUs like the M8s are far better for passenger service than locomotives pulling coaches: they have higher acceleration thanks to having their motors distributed along the train (distributed traction), and they are cheaper to operate. In fact, the M8 is the MTA’s fastest-accelerating [10], most reliable [11], and cheapest train[12]. What’s more, there are already a lot of them, with 471 built, including some specifically ordered for PSA [13], so there should be enough M8s for PSA service when it opens in 2027.
Looking at best practices from around the globe, almost all recent new orders for rapid transit (metro, commuter, or regional rail) rolling stock are EMUs. There are a smattering of battery, diesel, or hydrogen dual-mode multiple units (MUs), but outside of the US and Canada, to our knowledge, orders for new locomotives for rapid transit are exceedingly rare [14]. By choosing M8s for PSA, the MTA was planning on following the international consensus and giving its riders fast, frequent, and reliable service and delivering it as soon and as cheaply as possible [15].
Instead, the MTA has bewilderingly chosen to buy 13 new battery locomotives for PSA. This lacks sense on multiple fronts:
The entire PSA route is fully electrified, so there is no need for batteries.
If they need locomotives, the MTA could acquire spare locomotives from Amtrak, as Amtrak is expected to start replacing the Northeast Regional’s ACS-64 electric locomotives with Airos in the next few years [16].
If they needed locomotives because of the lack of catenary in West Side Yard, they could either put up wires on some yard tracks or through-run PSA trains to New Jersey, as we recommend as the first step in through-running in our prior work on commuter rail modernization.
At $23.45 million each, the battery-powered Siemens Chargers are, as far as we can tell, the world’s most expensive locomotives ever.
The battery Chargers will not be ready until 2029-2030 [17], while PSA is supposed to open in 2027-2028, two years earlier.
PSA trains will run frequently with short stopping distances in the Bronx, so fast acceleration is crucial. Locomotives, without distributed traction, have abysmal acceleration whether electric, diesel, or battery, which is why the industry has moved away from locomotives and towards multiple units.
The slow-accelerating locomotives will not only be slow themselves, but will slow down EMUs behind them and require increased schedule padding, slowing down the entire New Haven Line.
Metro-North plans to run 6–10 trains per hour in the peak direction and 4 trains per hour in the reverse-peak direction for PSA, meaning 13 locomotives will likely not be enough for full service throughout the multi-hour AM and PM rush hours.
Since the MTA does not have enough coaches, and the current ones are old and in poor shape, they also plan to order more coaches [18]. Concerningly, recent unpowered coach orders have been $203k/m [19], nearly twice as expensive as electric M8s at $112k/m.
If the intent is in part to use PSA slots on the New Haven Line for through service from the currently unelectrified Danbury and Waterbury branches, these branches could, based on federal estimates, be electrified for half the cost [20] of this order and achieve the same objectives by providing better quality service.
No agency has ordered new passenger battery locomotives before, as all battery passenger trains to-date have been battery EMUs (BEMUs) [21]. Thus, the MTA will be the guinea pig for an immature (and unnecessary) technology.
Almost no agencies outside of the US purchase new locomotives for rapid transit, yet the MTA is doing so now.
If the MTA wishes to run through service on unelectrified branches and does not wish to electrify them, BEMUs would be cheaper, not experimental, and accelerate on par with the M8s.
An Immature, Expensive Technology
Battery locomotives for passenger use are an immature technology even relative to other battery-electric trains. All battery passenger trains that ETA is aware of are multiple-unit sets. These trains almost entirely serve lightly-used lines that have little in common with the busy Northeast Corridor. This is important because BEMUs consistently have a twofold premium over standard EMUs [22], and as frequency increases, operators must run more, longer train sets. In Germany, where BEMUs have found the most use, they are deployed on regional lines connecting two suburbs without passing through the main city, running two- and three-car trains every hour or half hour. Lines running service akin to that planned for Penn Station Access were usually wired decades ago, and those that aren’t are being wired now.
But while BEMUs are at least a tested technology for appropriate use cases, battery locomotives as the MTA are proposing to use are, to our knowledge, never used for passenger service. They are absolutely not ready to pose as the standard for clean, innovative, and reliable transit that the rest of North America looks up to as the governor and MTA claim [23].
Crucially, the locomotives will use batteries for the last few congested miles into Penn Station, which is already electrified and where reliable operation is paramount. Ironically, Amtrak’s recent report, which erroneously claimed that through-running at Penn Station is impractical, nonetheless correctly cited exactly these avoidable mistakes with an interoperable fleet [24] and interoperable traction power systems [25] as key obstacles to instituting it.
Furthermore, even for battery locomotives, these battery Chargers are excessively expensive. For comparison, the MTA ordered 25 Wabtec R255 battery-diesel-electric locomotives for NYCT work trains for $233 million. At $9.32 million/locomotive, that’s almost 3x cheaper than the battery Chargers [26]. Union Pacific also bought 20 battery-diesel freight locomotives, 10 from Wabtec and 10 from Progress Rail (Caterpillar). The Wabtec locomotives have 2.5 MWh batteries. These were purchased for more than $100 million [27], at more than $5 million per locomotive, but still much cheaper than the battery Chargers.
In 2022, Metra tried to retrofit their diesel locomotives with batteries at a cost of $5.77 million/locomotive, but by 2023, was unable to come to a deal with Progress Rail, likely due to the immaturity of the technology. Instead, they have now ordered Stadler FLIRT BEMUs for the line.
Better Options
Besides the ongoing work to extend the third rail from Harold to Gate Interlocking and using M8s, there are multiple other options that the MTA should consider before buying battery locomotives.
The ideal long-term solution would be to procure a new EMU, the M10, capable of running on all types of electrification, as described in our recent piece on MTA commuter rail rolling stock procurement. EMUs accelerate and brake much more nimbly than locomotive-hauled sets, making them better suited to the demands of commuter rail service. The M10 would be similar to the current dual-voltage M8, but would instead be tri-voltage, adding a 25 Hz transformer. Worldwide examples, as well as the M8 itself [28], show that this would not come at a significant price premium. What’s more, procurement under the FRA’s new alternative compliance regulations would allow the purchase of lighter, cheaper, off-the-shelf international rolling stock conforming to European crash standards. With such lighter modern rolling stock, the added weight of a 25 Hz transformer should not be a problem. However, it is unlikely that an M10 could be procured and built within 2-3 years for the start of PSA service in 2027-2028.
A more immediate solution, besides the M8s, of course, would be to add an option to NJ Transit’s MultiLevel III order [29], which is capable of running on both 25 Hz and 60 Hz AC catenary [30]. And although locomotives are very ill-advised, the MTA could have bought non-battery locomotives like Amtrak’s ACS-64 or dual-mode and dual-voltage locomotives like NJ Transit’s ALP45-DP. They may have even been able to acquire the ACS-64s from Amtrak once the Airos arrive soon.
The MTA and governor explicitly said that the reason for the order is to run in and around Penn Station in battery mode, but on the off chance that what they meant was that the trains would run in battery mode on the unelectrified branches and in electric mode on the trunk, there are also better ways to achieve that. According to a recent FRA study, electrification should cost approximately $2.86 million per single-track mile [31], so it would cost $68.35 million to electrify the 23.9 miles of the single-track Danbury Branch, and $81.51 million to electrify the 28.5 miles of the single-track Waterbury Branch [32]. Electrifying both would only cost $150 million, less than half of the $305 million cost of the battery locomotives. Even if electrification costs were higher than the FRA estimate and very expensive like Caltrain, it would still be cheaper than the battery locomotives. Moreover, this would allow the existing M8 fleet to run on these branches, providing faster EMU service.
If full electrification is not feasible, even though it is the best option, BEMUs could be acquired for the unelectrified branches. Like in Europe and Japan, this uses BEMUs for infrequent branch service, not the core route around Penn Station. This would provide EMU-like acceleration, some compatibility with the existing EMU fleet, and higher but not absurd rolling stock costs, all of which would be far better than the battery locomotives. Such trains could either run through to Penn Station or require a transfer at the branch points, that is Stratford or Bridgeport for the Waterbury Branch and South Norwalk for the Danbury Branch; service today is a mix of those two options.
Conclusion
Transit planning isn't always a simple matter. There are many options to consider, and sometimes, finding the best solution requires counterintuitive or outside-the-box thinking. Often, however, the simplest thing is the right answer, and that is certainly the case when it comes to running simple electric trains along already electrified tracks. There is no need to buy expensive, exotic battery locomotives when the wires are already in place. It is simply common sense: normal EMUs are cheaper, faster, and more passenger-friendly. That’s why they are the standard across the country and the world.
Sometimes, when stuck in the trenches of a project, it can seem like an exotic, complicated solution is the only possible way forward, when in reality, it's clear from the outside that the simpler path is the correct one. That seems to be the case here: there is no world in which battery-powered locomotives are the right choice for Penn Station Access.
Purchasing these locomotives is a massively expensive—and frankly baffling—mistake. If at all possible, the MTA should convert this order to appropriate EMUs, if it needs new trains at all. The agency also needs to examine its decision-making processes. More often than not, the right answer will not be to blaze a new, untested, and expensive path, but to follow the best practices shown by New York’s peer cities around the world. Simpler is often better. The MTA should follow plain common sense and run electric trains on already electrified tracks.
Contributors
Blair Lorenzo
Vinay Madhugiri
William Meehan
Samuel Santaella
Khyber Sen
We wish to acknowledge the following ETA members who contributed to this report:
Ya Sin Allah
Madison Feinburg
Kara Fischer
Robert Hale
Tim Lazaroff
Alon Levy
Footnotes
[↑] Image: MTA, “The first Metro-North Railroad Siemens Charger SC-42 dual mode locomotive was unveiled at the railroad's Harmon shops on 2024-10-31. Picture shows locomotive #301,” (October 31, 2024, https://www.flickr.com/photos/mtaphotos/54110352409/), edited by ETA.
[↑] https://www.mta.info/press-release/icymi-governor-hochul-announces-metro-north-operate-first-nation-battery-and-electric
[↑] $304.9 million / 13 locomotives = $23.45 million/locomotive.
[↑] $112k/m * 85 ft * 8 = $23.21 million, where $112k/m is the cost/meter of M8s in 2023 dollars in the Transit Costs Project’s Rolling Stock Database.
[↑] MTA Board Member Veronica Vanterpool noted in November 2016 that part of the reason for acquiring the final tranche of M8 cars was Penn Station Access.
[↑] Currently, the New Haven Line and Hell Gate Line up until Gate Interlocking in Queens are electrified with overhead catenary wires carrying 12.5 kV AC electricity at 60 Hz. From Gate Interlocking to Harold Interlocking/Sunnyside Yard, the same style of overhead catenary wires instead carry 12 kV AC electricity at 25 Hz. From Sunnyside to Penn Station, there is both 12 kV 25 Hz AC catenary overhead and 750 V DC LIRR-style over-running third rail. Within Penn Station, all tracks have 12 kV 25 Hz AC catenary, and all but the southern stub-end tracks 1–4 (which are only accessible from New Jersey) have LIRR-type third rail. The West Side Yard only has LIRR-type third rail.
[↑] The MTA has released recent PSA project updates on beginning construction for DC (third rail) substations HG-01 and HG-02 on the Hell Gate Line in Queens, started in Q1, 2024.
[↑] We note in our recent piece The Future is Electric that the LIRR’s recent capital plans for third rail expansion have been between $18 and $32 million per route mile, compared to only $3.2 million per double-track mile for AC catenary according to a recent FRA electrification study from January, 2025.
[↑] The MTA’s subways and EMUs have the same initial acceleration of 2.5 mph/s, but due to its relatively lighter weight (only compared to the M7 and M9A; it is still quite heavy) and 12.5 kV AC power, it has the best sustained acceleration curve.
[↑] At around 800k miles, the M8 has the highest MDBF of all MTA rolling stock.
[↑] At an internationally-respectable price of $112k/m, the M8 is cheaper than any recent MTA train order, including subways, M7, M9A, coach cars, or locomotives.
[↑] MTA Board Member Veronica Vanterpool noted in November 2016 that part of the reason for acquiring the final tranche of M8 cars was Penn Station Access.
[↑] In the US, the MTA in NY, Altamont Corridor Express (ACE) in California, Trinity Railway Express (TRE) in Texas, and MARC in Maryland have all bought Siemens Chargers locomotives. In Canada, Exo has also bought Siemens Chargers. Outside of the US and Canada, passenger locomotive orders have only been rebuilds, such as Peru buying retired F40 locomotives from Caltrain or Lagos buying unused Talgo locomotives from Wisconsin. Almost all other new locomotive orders appear to be for intercity service or freight. The few exceptions appear to be Turkey’s TCDD E68000s, used for some regional services and the Philippines’s INKA CC300s, used for some commuter services.
[↑] At least concerning rolling stock, as there have been delays and cost overruns for other PSA capital work.
[↑] The most recent single level coach order in the New York area was by CTDOT for an Alstom 60-car order for $315 million: $315 million / (60 * 85 ft) = $203k/m (ordered in 2023).
[↑] $2.86 million/single-track mile * (23.9 miles single-track Danbury Branch + 28.5 miles single-track Waterbury Branch) = $150 million < $305 million.
[↑] There have been new freight battery locomotives, but no new battery locomotives for passenger service as far as we are aware. The locomotive-hauled, high-speed TGV M does have backup batteries, but these are backups in case of emergency, capable of only running at slow speed for a short distance, not in regular service. Metra also considered retrofitting 6 diesel locomotives with batteries, $5.7 million per retrofit for the Rock Island Line, but could not reach a deal with Progress Rail, realized their error, and has instead purchased Stadler FLIRT BEMUs for the Rock Island Line.
[↑] Stadler has sold FLIRT 3XL EMUs in Bremen for $104,000/m: €100 million / (16 * 87 m) / 0.69 $/€ = $104k/m (ordered in 2022). Hanover bought Stadler FLIRT 3XL EMUs for $110,000/m: €320 million / (64 * 68 m) / 0.69 $/€ * 1.03 = $110k/m (ordered in 2018, manufactured between 2020-2024). Austrian FLIRT Akku BEMUs are $249,000/m: €1.3 billion / (120 * 63 m) / 0.69 $/€ = $249k/m (ordered in 2023). Chicago’s Metra’s FLIRT BEMUs are $197,000/m. The small combined Metra order (8 2-car BEMUs base order, 8 2-car BEMUs with 32 extra cars for the option) amounts to ($154 million + $181.4 million) / ((8 * 2 + 8 * 2 + 32) * 51.5 m / 2) / 1.034 = $197k/m (ordered in 2024). However, the addition of the 32 unpowered cars makes these BEMUs quite underpowered.
[↑] https://www.governor.ny.gov/news/governor-hochul-announces-metro-north-operate-first-nation-battery-and-electric-powered
[↑] https://pennstationcomplex.info/wp-content/uploads/2024/10/Doubling-of-Trans-Hudson-Train-Capacity-at-Penn-Station.pdf#page=164
[↑] https://pennstationcomplex.info/wp-content/uploads/2024/10/Doubling-of-Trans-Hudson-Train-Capacity-at-Penn-Station.pdf#page=164-165
[↑] The Chargers can go faster, but don’t have diesel engines as well like the R255 (and have pantographs instead of third rail). The R255s have 675 hp Cummins diesel engines and a 500 kWh battery. These were ordered in 2020 and delivered in May-June 2024, entered service in September 2024, and were certified in January 2025.
[↑] Unfortunately, Union Pacific does not give a precise cost and only says more than $100 million.
[↑] The dual-voltage M8 is already cheaper ($112k/m) than the single-voltage M7 ($121k/m) and M9 ($163k/m).
[↑] The MultiLevel III is an EMU comprising powered and unpowered cars.
[↑] https://www.njtransit.com/press-releases/nj-transit-board-approves-purchase-new-rail-cars
[↑] The FRA study found the cost of overhead catenary to be $2.86 million/single-track mile and $3.2 million/double-track mile.
[↑] Costs for electrification of the Danbury and Waterbury branches would need to be paid for by CTDOT, and should the Springfield line (which the Hartford line runs on) be electrified, too, this would need to be paid for by Amtrak, CTDOT, and MassDOT. Electrifying the 60.5 mile line would cost approximately $193.6 million. We have included this number here as well as it would be the only remaining diesel branch supported by CTDOT at that point.