Hyundai Nexo sets new world record 887.5 km travelled on single tank of hydrogen

NB that I'm an electrical engineer and have built many batteries before, including from 18650s. First off, it is absolutely not necessary to have so many controllers; cell-level controllers are only used in the cheapest batteries. High-quality batteries will match cells individually and switch them in groups, which reduces the losses from running a bunch of parallel circuitry. This is not really different from simply using larger cells, as long as you can match them sufficiently well.

Cell-level BMS is very cheap as a result, for example this single cell chip that costs 7 cents each: https://lcsc.com/product-detail/PMIC-Battery-Management_HY22...

> Plus the hassle and cost of using bus bars instead of wires.

Only if you're using 18650s; if you are using prismatic cells you just solder or crimp the tabs to the HV supply. If you're using small-format cells, you would typically used stamped sheet current carriers rather than busbars.

> And it means that the battery module output is 3.6V at an insane amperage and needs a massive transformer for a reasonable voltage.

It certainly does not, because batteries are DC and you do not use transformers to convert their voltage. Transformers require you to convert to AC and back, but you can use an inductor ~100x smaller by using one of the many "boost" topologies. The basic idea is that you only need to convert to AC, and you don't need to rectify back to DC. You chop the high-current DC up very finely, then store 1-3 cycles in a very small inductor. If the inductor is sized just right, it will drain out current that is very close to DC.

Naturally though you would really want to operate between the 245 volts of the Mirai's battery or the 650 volts of the Mirai's fuel stack. That would be something between a 68s27p pack and a 180s10p pack. That's far more efficient. You can still have cell-level BMS for every single cell- that IC I linked above is daisy-chainable, meaning it does not need a real 0v ground reference. It's perfectly happy working at whatever voltages on the two ends of the particular cell its connected to. In this case I'd opt for something closer to 68s- the more cells you have in series, the more susceptible your design is to degraded performance over time.

For trucks where you need bigger tanks, both autonomy and boil-off rate improve by themselves due to the square-cube-law. And you can use the boil-off to charge the batteries in your hybrid electric powertrain, run auxiliary systems etc.

The LH2 trucks currently under planning have sophisticated fuel management systems where the driver inputs where she will stop and for how long, and the truck will optimize LH2 consumption and battery SoC before arriving at the stop point.

As for LH2 versus compressed: do you actually know how energy demanding a hydrogen compression train is? Compressing a low molecular weight gas to 700 bar is just fundamentally inefficient.

At the scale required for mass adoption of hydrogen fuels, both compression and liquefaction are talking 6-8 kWh/kg. For comparison the energy content in hydrogen itself is 33 kWh/kg. So naturally there is great interest in recovering as much as possible of that work before consumption.

- Transport. You can either transport hydrogen via pipelines or simply produce it via electrolysis in-place. You have to get the electricity to the destination somehow, so this doesn't really differ between electric and hydrogen cars.

- Building up pressure. Energy required to build up pressure is related to how much the volume is reduced (force x distance and all that...) Conveniently water is almost incompressible, so you can raise the water to the desired pressure cheaply before you electrolyse it. https://en.wikipedia.org/wiki/High-pressure_electrolysis says this can be done at 97% efficiency.

Hydrogen is an essential industrial input and feedstock for many processes, including production of fertilizer in the form of ammonia. So we will be producing hydrogen for that, without a doubt.

It is estimated that by 2030 or so that electrolyzers will be as cheap as steam methane reformation (SMR), with or without carbon capture on the SMR. We will see if that prediction pans out! The IEA has put approximately 850GW of hydrogen production as a feasibly goal by 2030 as part of their plan for keeping climate warming at 1.5C, with half of that from electrolyzers. This may be fairly plausible, as Europe has mandated 40GW of electrolyzers by 2030, and there are currently 154GW of hydrogen electrolyzers in the pipeline:

https://www.rechargenews.com/energy-transition/growing-ambit...