HN Theater @HNTheaterMonth

Batteries are now a viable option. They can pull power from the overhead lines or third rail when available and run on batteries when power isn't available. The LiFePO4 batteries are a drop-in replacement for existing diesel generators, so producing them is easy.

https://www.youtube.com/watch?v=9s4heZe7ChM

⬐ stickfigure

For a single small passenger car without any grade, maybe.

From that video: 90kw/hr of battery. At peak power output of 180kw, you could run that for 30 minutes.

The ubiquitous GE Dash-9 used in freight service here in the US generates >3000kw and can run all day. They're also usually run in multiples to pull long heavy trains.

I wouldn't expect to see battery powered trains in the US anytime soon, except maybe short passenger-only lines on the east coast (which are probably viable for electrification anyway).

⬐ Robotbeat

You, strangely, ignored the parent's point that you would just be using the batteries to reduce the amount of overhead wires you need, i.e. one would have opportunity to recharge several times a day.

(It's not helpful to focus on the specifics of a proof of concept--unimportant in this discussion about what is feasible--versus the fundamental figures of merit of the technology... Unless one is prepared to make the case that the specifics are the limits of the technology.)

There are 100MW power, 100MWh energy lithium ion batteries installed already (with 1GWh on the way). There's no technical reason you couldn't have battery powered trains. The "batteries are too weak" argument is dead.

⬐ creato

Can you even transfer close to 3000 kW through any overhead wires or third rail systems? It seems like 25 kV is the max voltage overhead lines run at. That's 120 A that you need to transfer through a sliding conductor. That sounds problematic to me, but I don't really know for sure...?

And if you are proposing using batteries to reduce the amount of overhead wires you need, then the average power you need to transfer while connected to the wires increases, not decreases.

⬐ hvidgaard

If we take the hypothetical electric locomotive with a battery pack, it would probably decrease the peak load on the lines. You only need maximum power at acceleration and climb. The battery is perfectly capable at helping with the peaks, while recharging during the cruise when the power demand is low.

⬐ iggldiggl

To give some ball park figures, this is the default maximum current permitted per train according to national regulations:

- UK conventional lines: 300 A @ 25 kV

- UK third rail network: 6800 A @ 750 V

- German conventional lines: 600 A @ 15 kV

- German high speed lines: 1500 A @ 15 kV

⬐ cardinalfang

A TGV can take over 12MW. It uses a single contact feeding the two power cars at each end of the train to avoid problems with oscillations in the overhead line at high speeds.

⬐ TylerE

The most powerful locomotives are all electric. Powerful freight locos have power outputs of over 13000kw

⬐ stickfigure

I don't think you've run the numbers.

A tesla powerpack is 50kw, 210kwh, and weighs 3500lbs.

A freight train that would normally be pulled by four 3,000kw diesel electric locomotives would need 240 powerpacks, and could run for four hours before needing a full recharge. Those powerpacks would weigh 420 tons - the equivalent of say, four fully loaded freight cars, and cost around $25M (plus whatever the locomotives cost).

The diesel-electric locomotives are a couple million each, ready-to-run.

Ah hah you say, you only need the batteries between catenaries! How fast do you think you can charge those batteries? If you can charge the same as the full discharge rate, then you need 50% of your track to be electrified, and you need it every four hours (assuming you're willing to risk full discharge cycles). Trains don't move very fast, so that's pretty closely spaced. And even worse, you now need electrical infrastructure with twice the capacity - you need to charge the battery and move the train.

Sure there's no technical reason you couldn't do this, but the economics are not looking good. Batteries are too weak.

⬐ kzrdude

420 tons of batteries sound like an insane resource consumption in terms or metals and rare earth metals.

⬐ Robotbeat

You just did it again. Used specifics from one application to pessimistically (and wrongly) apply to the technology generally. Also, you keep assuming the 4 3000kW locomotives will be running flat out, which is a terrible assumption (and would cause a conventional locomotive to quickly deplete its fuel, if not destroy its engine).

Here, I'll do it for you. Good Panasonic cells get about 250Wh/kg, or 0.9MJ/kg. Assume an electric-optimized locomotive would be able to achieve about half its weight in cells, with a useful energy density of 0.45MJ/kg. Assume about 1 locomotive for every 9 cars, and with each car weighing the same as each locomotive. So the whole train's effective energy density is 0.045MJ/kg.

The "rolling resistance" of a typical train is about 0.002, conservatively. That is a weight of 1 kgf has a resistance of 0.002kgf. (EDIT: This is a good assumption that works up to 60mph, the speed limit of freight trains, but at the typical low average speed of freight trains, it's actually about half that value: https://slideplayer.com/slide/4696076/15/images/12/Freight+T... )

The range is thus just: (specific energy)/((rolling resistance) * gravity) or: 0.045MJ/kg/(.002 * 9.8m/s^2) = ~2300km. https://www.google.com/search?q=0.045MJ%2Fkg%2F.002/(9.8m%2F...

That's enough to go from the center of the continental US to the coast on a single charge. (and from what I understand, 1 engine for every 9 cars is not uncommon)

If we have one engine for every 4 cars, you can now cross the continental US on a single charge. But remember, the discussion was about multiple recharges per trip, so there's WAY more battery here than you actually need.

And to just give an idea of the power available, 130 tons is a typical car laden weight. 65 tons of Panasonic cells gives you 16.25MWh of storage. Cells like that can discharge their cells about 12 minutes. Lets make it 30 minutes, conservatively. That gives a power at the cell level of 32500 kilowatts, ten times your 3000kW locomotive. Batteries are plenty powerful.

(And the cost is offset easily in fuel costs, as long as the battery is given the usage of about one full cycle at least once a week.)

You might point out energy requirements for braking and climbing hills. But remember that one of the greatest advantages of battery-electrics is regenerative braking. Most of the energy consumed in increasing elevation can be recovered on the way back down.

(As far as costs go, the battery pack should--including the price of industrial electricity and typical costs for automotive batteries at scale--pay for itself in fuel cost savings in about 500 cycle-equivalents while the cells should last at least 1000... meaning the overall added cost is potentially negative... meaning it's a market opportunity.)

⬐ daurnimator

Why assume 4 hours with a recharge? I'm thinking closer to max 15 mins without a recharge. This would allow skipping electrification in some tunnels; in rail yards; on track segments shared with non-electrified trains; and anywhere that it's too hard to run electric lines.

⬐ Gibbon1

Far as I can tell he's also off by a factor of 4.

Tesla 85hw battery pack 1200 lbs.

3000 kw X 4 hours = 1200 kwh.

1200kwh/85kw*1200 = 170,000 lbs.

A GE/NS Dash 8 weighs in at about 390,000lbs

⬐ stickfigure

210kwh for 3575 lbs comes from the tesla powepack specs:

https://www.tesla.com/powerpack

Furthermore, it makes zero sense to compare "3000kw x 4 hours" since diesel electric locomotives can run for N>>>4 hours.

Furthermore, you're comparing the weight of batteries alone compared to the weight of a whole locomotive.

⬐ Gibbon1

> Furthermore, it makes zero sense to compare "3000kw x 4 hours" since diesel electric locomotives can run for N>>>4 hours.

This is simply you reframing the conversation from using batteries to allow hybrid electric trains to use short sections of non-electrified tracks to hybrid electric trains won't work because they don't have the range of a diesel locomotive.

I'm going to put this down as you're unwilling to argue fairly and thus lost this argument.

⬐ stickfigure

You're not comparing hybrid against diesel-electric, you're comparing hybrid against fully electric. The cost of putting batteries in your rolling stock might or might not be higher than electrifying the last 5%. Either way, it's small compared to the cost of electrifying the 95%.

Note that the eastern seaboard (with its bridges and tunnels and topography) is getting electrified; the long and flat midwest is not.

⬐ stickfigure

Sure... but now you're only saving a small percentage of the total electrification cost, and paying for increased cost, complexity, and maintenance of locomotives. It's not an obvious win.