The issue with large batteries for locomotives or OTR semi's, is the large current draw to recharge them in a reasonable timeframe. I would think that the batteries and the vehicle could be designed so that the batteries could be lifted out and a newly charged battery reinstalled. Of course that entails a large capital investment.

Why did Kennecott Copper have battery locomotives? Because track in copper pits tends to get pushed around to get to the ore, faster than the trolley guys can keep up.

So for UP, many reasons. #1 political/optics, it makes it look like UPRR is giving it an honest try.

#2 learning. Bob Lutz didn’t just wake up one day and say “Imma make a million Chevy Bolts”. Had to play with conversions, do the EV1, 2004 Silverado hybrid, Volt (first hybrid that could turn around at the end of my court without the engine running) and then come round to the Bolt. So this os partly “cutting teeth” i.e. to start to develop competency.



Alternative fuels? LOL Yeah, the problem is there are way too many ideas that are very costly to build in small quantity. You don’t need to fight through that steep learning curve when you use established COTS stuff. Lithium batteries are already being made at scale.



Well. There’s the real nut of the matter. Someone, Dave Johnston maybe? explained to me the industry’s apprehension about electrification. A huge problem is grid prioritization - railroads are afraid some goody-two-shoes grid operator would have a bad day, need a big load to cut, and cut the railroads because it’s easy. Or they don’t want their operational security (e.g. hacking attacks) dependent on the chaotic mess of players in the energy market.

Now, as to grid stability, we’re entering an absolutely huge “learning experience” as we populate the nation with electric cars capable of either disrupting the grid, or *stabilizing* the grid, depending on how skillfully they are used. Alternative power is injecting power into the grid at odd times, so energy storage is hugely incentivized. And 1980s era microcontroller tech is finally making it into energy management - e.g. the SPAN panel for homes. We’ll simply have to see how all this shakes out - if it’s done as a masterwork of competence, it should “put to rest” any fear of grid stability. In fact, after winter 2023 Ukraine, countries may consider it a national-defense priority.

Another problem is the sheer difficulty of running 11,000 volt wires inside tunnels. But batteries are a game-changer here because they let you do Detroit M-1 Rail style installations where you just delete the overhead anywhere it’s inconvenient.



Spending *that kind of* capital on national infrastructure is not the role of a private company. It’s the role of a government, at least that wants to be *in any way* competitive in 21st century economies. I know some people who hate socialism and I know others who hate militaries. Unfortunately YOU don’t get to decide how much you need — your adversaries decide that. You can, of course, choose to be conquered or left behind in the world economy.



That sounds like anti-electric political faff honestly. Now that we have some experience with them, most EV batteries are lasting the life of the vehicle, with (judging from the market price of used EV batteries) a very vivacious after-life.




Words cannot express how disappointed I am with that. “To nowhere” is because they are doing the easy parts first. The entire mindset is to trick the state’s entire population into falling into “the fallacy of sunk costs”, now that we built 10% we must spend 20 times that to finish it. It’s a marketing scheme. It may not work.

For cheaper than what they’ve spent, they should’ve started at the pinch-point for the whole enterprise - the Grapevine Base Tunnel (paralleling I-5, the straight shot from the Central Valley to L.A.) They’re skipping that route because it’s so very costly. But if they’d started there, freights and Amtraks would be using it the day it opened, so you’d have this beautiful revenue stream to help stand it up4, and then people would *practically demand* the easy parts be finished.



They surely have, Boy Howdy… and the prices have dropped like a stone. I was shopping for used Nissan Leaf batteries and came upon a Nissan dealer selling new ones. 62 kWH for $13,500 with core ($15,500 without). I know that sounds like a lot, but think - That’s 25 cents a watt-hour! Dealer price! Used, Tesla Model S packs (6 kWH) used to be $1200 and are trending now toward $800 or about 13 cents a watt-hour.



The 1990s called, and want your political talking points back lol. All this stuff is ancient history. The batteries work and EVs are over the subsidy hump.

I was rooting for EVs and it’s going *much, much* better than I expected. I never imagined early EVs would be good for 300 miles (I didn’t imagine battery density would get there in my lifetime). I had no idea how ridiculously easy home charging is without a service upgrade (really) or major grid improvements. It’s all about *time* and a little bit of 1990s era microcontroller tech and tiny amounts of datacomm so sort out the “when”.

Of course automakers can tell people to like *ANYTHING* with enough advertisements - that’s why anyone wants an SUV instead of a station wagon.

I was raised on the idea of Progress, an unflinching optimistic hope for a better world and a plan to make it. These are core 1960s values. I find it odd that people who want to live in the 1960s do not share its most basic value :)




Quite right. Nothing prevents adding (or swapping) battery tenders.




I have my doubts that they will bother developing any radically new battery tech for trains. Lithium production has already scaled up, and developing something radical with the big engineering curve doesn't make much sense when there's a COTS thing you can just buy that's already cheap because of economies of scale.




This is the root problem with low-volume production and radical tech. The Green Goat predates our broad experience with battery EVs, so they were on the bleeding edge getting cut while Tesla learned from their mistakes. That's why you stick with established tech.

If I was making a branch/ switching locomotive, I wouldn't try to invent anything new, I would just collect 100 Tesla Model S battery packs from wrecks. It's an extremely well-understood platform built at scale. So I get to apply all that invaluable learned experience. It's easy to just program soft limits that gate the crews from destroying the battery with bad decision making.

Trip planning is also easy. Anyone who's given EVs a chance knows the trip computers are quite good, and help a great deal with trip planning. So the same thing can be done with trains. So management has a viable trip plan before the crew is even called, no need to send the machine on suicide missions.

Yeah, 100 Model S packs, 8 megawatt hours or a 3000hp engine running wide open for 4 hours, which is not how branch locomotives are actually used. Not quite ready for Donner Pass, *but not as far away as I would have assumed*.

I would absolute use packs from cars straight-up, because of the absolutely amazing internal design, packaging, cooling, onboard management systems and that they have an internal contactor. That means you don't need an external one. It's just not worth trying to develop something original for a continental market of perhaps 5000 units.

In one of our technical presentations here at the university, one point made was that most of the European electric network is three phase power. The homes in Copenhagen have three phase. This means car charging is much faster in Denmark than the equivalent single phase American home.

Meanwhile, construction is nearly finished and trains have been delivered to western Denmark for conversion to battery power.

https://mjba.dk/danmarks-foerste-batteritog-rammer-skinnerne-i-vestjylland-i-2024/

https://www.vestkysten.nu/batteritog-paa-vej-til-vestjyske-skinner.html

Getting back to preservation, I could see an application for this on for example Mt Washington Cog. They have long layovers at base where charging is possible, and strong regeneration options on the return journey.

Battery powered Electric Vehicles (EV) are a dead end technology.

A senior executive from Ford Motor Company ( a manufacturer with a century of proven execution supplying USEFUL and COST EFFECTIVE motor vehicles ) recently stated that EV's would not be "cost competitive" against ICE vehicles until 2030 AT THE EARLIEST..

Rivan a New "startup" in the EV marketplace is about to be "delisted" from NASDAQ because they cannot produce a product that people will willingly purchase at the production cost (with a profit).

The EV dream is headed to the scrap heap of history AGAIN...

The solution is no different from what it was in the 1970s: use parallel or massively parallel charging and divide the crossbar switching of the batter subcell arrangement so that you're using a manageable current per section. That is realizable when capacity of the power consists allows no more than a certain number of discrete charging points (with the same operational model as diesel refueling often is) and adequate emergency 'rescue' charging equipment is made appropriately mobile (something that even defectively-implemented level 4 autonomy facilitates, but that's another story).

Another improvement in charging time has come about more recently, with the costing-down of effective super/ultracapacitors. These can source very high currents (albeit at low voltage per device) and can then optimally 'charge' the chemical cells to allow another locomotive to gang-connect.

The concern I have is that I don't think the industry quite learned the lesson of why the Green Goat was such an abject failure. It seems as if they thought 'switching' was a relatively easy service compared to Getting Trains Over The Road, whereas in reality it was one of the most demanding cycling for a chemical battery you could find -- and that was before PSR put so much increased reliance on rapid flat switching of blocks. I *think* RPS understands the situation well enough, but I don't see other architectures with quite the robustness that even cheap locomotives expected to be switchers would need to have.

With all due respect, speaking as a degreed electrical engineer with over 4 decades of experience "pushing electrons around" it is not any easier with "super capacitors" in the mix.

Sure, a super cap can spit out a boatful of electrons "on demand" but you still have to "fill" the super cap with a boatful of electrons before it can "spit" them out "on demand"....

There are no "short cuts" when it comes to packing a chit ton of electrons into a chit ton of chemicals.... Inject the electrons really fast and overheat the chemicals or slowly trickle the electrons into the pile of chemicals and wait two days before you can move your locomotive again...

These are all well understood laws of physics, the sooner you accept them the sooner you will realize that a practical EV Railroad Locomotive is a pipe dream....

Henceforth this is why for the logistics industry EVs will be nothing more than something bought to keep the environmental wackos happy. Nothing on an effective basis will replace the simplicity of carrying diesel fuel and burning it to provide the needed power to move the goods. California can demand all they want but when reality bites them in the rear end alongside a few kicks in their nuts and face they will have to accept the facts. But this is California we're talking about the same state that thinks 80 billion dollars in free money to a certain group is a good policy when they are facing a 32 billion dollar shortage in their own budget.

A more current (no pun intended) EE would surely recognize that the supercap charge rate is comparable to other capacitors in being very rapid, and having limited damage from either high charge or discharge rates -- unlike most forms of chemical battery. In this application the supercap array acts as a 'charge buffer' to accommodate the current at the rate it needs to be sourced, and then release that under control at the desired charge rate of the chemical storage.

As a member of IEEE for over a quarter century, I stand by the soundness of this approach.

I also stand by the development of AC-drive dual-mode-lite combined with battery/supercap storage (whether in a dedicated unit like a FLXdrive or Mr. Iden's power tenders) as an improvement on current straight diesel-electric power. GE expended a great deal of effort on a 'hybrid' locomotive around the time of the 2008 financial collapse, but they were trying to integrate the traction battery on the locomotive itself, and only as automobile-hybrid-type regenerative or engine-driven charging. While I think that option remains technically feasible, there is little point in building new (and obligate tier 4 or 5-compatible) hybrid locomotives without electrification pickups or the ability to share power and DB, rather than arrange to have any fancy new green power be quickly usable with existing power, and built on a common cabbed and traction-motored frame from "non-zero-carbon" locomotives.

Every pole top in Europe (even Russia) has the same: 3-phase "wye", 230V leg-neutral and 400V leg-leg.

3-phase sounds good, but e.g. a typical modern house gets 240V@200A (48kW) vs a large German home getting 63A @ 230V 3-phase (44kW). It's actually less power.

But it doesn't really make a difference to cars. Even 11kW charging is "bonkers" overkill for the needs of daily commuting for most people. I help about 3 people a week successfully setup EV charging at home, and much of that is "walking people back" from wanting HUGE charging that their service cannot support. NEC 220.82 is a buzzkill :) See this video; fast version at 32:55).

There's another hard limit. The "charger" is just a gateway, the actual AC charger is on the car and it costs money. Most cars have a 6.6 to 11.5kW onboard charger. This is trending downward as cars become more affordable and as manufacturers collect data showing higher power rates are rarely used productively.


Yeah. 4 years ago I asked for some help doing a number-crunch on that, and it looked feasible.
viewtopic.php?f=1&t=44068


Well, what they do with EV *cars, today* is place the batteries in a series string up to 400-800V, and then, massively parallel the chargers. For instance one company makes DC fast chargers anywhere from 10kW to 300kW... the 20kW unit has 2 modules and the 300kW has thirty. Cheap, swappable, losing 1 doesn't degrade the charger much.


Wait. Is “rescue refueling” currently a regular thing for locomotives? Do commercial jetliners regularly crash into the Pacific because nobody checked the gas? How do EV cars handle this today? (Hint: the car’s nav system is pretty assertive).

Because that doesn’t seem to be a problem for anyone today. Trip planning works, and we’re not going to stop doing it at the first sight of a battery. Much the opposite, the computer can see the fuel gauge, and will tell dispatch by commlink, so everybody knows at the outset if a planned trip will work or not. Probably even pop up on the train dispatcher’s display as an alert if a train is marginal. The computer knows the route grades and curves. Even if the crew doesn’t program the train weight, the computer will infer it within a mile and tell the crew “cannot make next yard”. So a bunch of people will get called on the carpet if it happens.

Realistically if a battery unit dies on the road, it needs to go to the shop.


Really? COZ WE’RE DOIN IT.

It’s live, in production, at scale. There are many 200,000 mile cars and a few 400,000ers that almost exclusively Supercharge (how else do you get 400,000 miles on a Tesla). All these cars “phone home” so the manufacturers harvest all this data, and Tesla shares to improve the industry. There’s a “chit ton” of hard data to be read. Successful engineers love hard data lol. Go look :)

So yeah, we know exactly how long it’ll take to recharge the locomotive. 18 minutes from 10-80% if you use Hyundai packs.

Readers ask how? Most EV batteries have about 2 gallons of coolant. Here’s the Model S loop that touches all 7000 cells. Bolt uses pouch cells, and cools the edge with cooling plates, but they super-chill the coolant with a heat pump. Leaf uses only “skin of the battery case” cooling, and that limits DC fast charging on road trips.

Actually, DC fast charging is now limited by the cable of all things. A water-cooled cable over 400A is just too unwieldy for consumers. Voltage is good to 1000, so the series voltage of the battery is a big factor in DC fast charge speed. That’s why this Hyundai can do 240kW (over half again Tesla’s ~140kW max) but can’t hit charger limit of 350kW.


Not so well understood, clearly :)

Yeah, iterative design can do wonders. It’s called “actually trying” and it’s kind of Elon Musk’s thing. Nobody in 1997 would've said Internet payments could be safe. Nobody in 2004 said you could have a 2-stage-to-orbit rocket and have the first stage come back and land at the launch site and be reused in a week for less than $1 million. Musk is like "challenge accepted" and then does it.

Just how big and heavy would the battery pack be for a 4000 hp locomotive? How long would it take to charge it and at what KW rate? And how far would said locomotive go before needing to be recharged? How much would a multi-locomotive charging station cost serving X number of charges per day? How many more locomotives would a railroad need compared to the number of diesels?

Really ? Where exactly are "WE" doing it where an EV (battery) railroad locomotive with a nominal horsepower of say 4000 hp is routinely operated on "regular" freight train schedules....

Looking for evidence....

Obviously you'll never compete with the energy density of a 5000 gallon diesel tank. Question is, *do you have to?*

Mind you, I'm not trying to fail, so I might look at it differently than others. I originally spitballed 100 of the 100kWh Tesla Model S packs, I feel comfortable with that as a starting point. You can hit the web or Youtube for how Teslas work and just move the decimal place over 2 places. You're an engineer.

Now, one refinement I'd make is to start with an SD40 and break it up into six individual 1-motor "locomotives" of 500 horsepower each. That is a much more practical size to think about. Tesla cars *already are* 500 horsepower. We're well within the range of proven EV products like the ZILLA and it's more comparable to work like transit buses, trucks, etc. You can draw on their experience.

But generally, charging isn't a race... unless it is. You want to charge at the slowest possible speed consistent with the mission. At "slow" charging comparable to EV destination charging, your 10MWH pack would need 1MW, and the regular old Tesla NACS can do that if you water cool everything and don't mind a cable the size of your wrist. There's also MCS. But honestly to go faster than that, I would just use multiple MCS or even NACS instead of trying to come up with a 10MW connector.


Thinking about thermal content of fuel and conversion efficiency of an engine (not good), I'd consider a 100kWH pack to be comparable to 10-20 gallons of fuel depending on hard-to-analyze stuff. So 100 packs = equivalent of 1000-2000 gallon tank. Probably not for every job on the railroad, but that's what trolley is for.



You mean like a trolley substation?

Interesting but I have some apprehensions about taking 25kV off a wire and using that to fast-charge. The reason is the DC fast charging equipment is bulky and heavy, and I'm starting to worry about weight and space after adding 100 Tesla packs :) It would certainly be elegant if you could squeeze it on board, because then, you could fast-charge while underway. Maybe it would be worth sacrificing some packs to make room for DCFC equipment.

Wow, thanks, this really is designing itself.

Anyway, costing is a big can of worms, and the elephant in the room is you're paying 12 cents a kWH instead of 30 cents to $1.50 a KWH to make it from diesel. Same basic economics of buying power from the grid vs making it at home with a generator.

Anyway if you aren't charging-on-board and you need an external DC fast charge station, couple million. They're building them like crazy so they're becoming commodified.


Hard to say. I think where you're going with that is it'll be spending hours or days on the charging rack; that ain't so unless you don't need the unit and have the time to spare. A lot of locomotives do sit around. You could charge it in 15 minutes if you really wanted to, or charge it enroute with trolley in strategic places.

Probably spend less time in the mechanical shop honestly. With 100 packs that's a lot of surface area for failure so you'll be swapping a few every year. If you stagger-stack Tesla packs, pulling one could be 20 minutes diagnosing and 10 minutes swapping, send it out into the consumer rebuild market for maybe $5000. Compare that to swapping a power assembly or turbo, you tell me.

You could develop a custom-for-railroad 100kWh pack if you really want to, and then it'd be $50,000 to rebuild. That's why I'm all about the consumer stuff. The great Google search engine was built on huge arrays of dirt cheap PCs.

I’m not going to get in the middle of this, however when I was a teenager I realized that there were only two kinds of people in the world.
“ Those that say it can’t be done, and those that find a way to do it”.
I’ve never had much time for the first kind!