Hacker News Reader

Comment by pg_bot | Hacker News Reader

You would need several scientific breakthroughs to make hydrogen economically viable for transportation. It's expensive to generate, it embrittles storage containers, is highly flammable, and there's no existing infrastructure that you can piggy back off of. Right now it is a dead end.

View on HN

Only 8500 coal power plants in the world producing 20% of emissions. No miracles needed to replace them.

https://en.m.wikipedia.org/wiki/Coal-fired_power_station

Replacing a billion gasoline cars is going to take decades

We should start by not making new gasoline cars, or new coal power stations

Electric cars aren’t ready. And absolutely everyone knows that, so why are we bringing it up?

We could’ve replaced coal plants two decades ago, however, and reduced emissions by 20% globally

> Electric cars aren’t ready

The scaling problems facing battery electrics are dwarfed by those facing hydrogen. They’re also being aggressively solved in America, the EU and China.

>They’re also being aggressively solved in America, the EU and China.

Challenge. Where and how are these problems being addressed, other than hoping that people voluntary reduce demand when asked?

Not sure what we are talking about.

There are an absolute shit ton of battery factories being build. A huge ramp up in science and research.

There is also a huge ramp up in cathode and anode production facilities.

All major lithium companies have announced major build out of existing facilities and building new once as well.

A major expansion of both natural and synthetic graphite is being done.

Cobalt is not expanding as quickly but its getting systematically kicked out from the batteries anyway.

In addition to that, LFP batteries reduced the projected demand for nickel from batteries.

In addition to that CATL is introducing Sodium batteries as well.

So maybe you were not paying attention but there is a gigantic explosion in every part of the battery supply chain happening right now. With major private and public investment. There will still be some shortages of important materials, specially lithium however its certainty being worked on.

Of course there are also improvements in safety, reliability and so on. A huge amount of research and development.

The gating factor for the scaling issue isn't the battery production (mining externalities notwithstanding), but the infrastructure to support them.

Wave your magic wand to will everyone a Tesla, and watch the grid collapse the same instant.

The grid has plenty of spare room in the evening "bath tub" (daytime peak is very different from nighttime peak). If you wave that magic wand and everyone has a Tesla today and everyone charges at wall outlet at off-peak times, the grid wouldn't hardly notice.

The grid operators are not stupid, they are also building generation to handle this.

While it remains to be seen if they build enough, they are planning for EVs. Where local laws allow it anyway. They see big $$$ from EVs replacing gas, and they want that money.

Yeah but nobody is waving a magic won, as production of EV will go up, ways in the network would be found to handle this.

Many studies have been done and this and every study I have seen agrees that its not as big of a problem as many make it out to be.

In fact, I assume infrastructure for charging for EV owners without a home is the much larger infrastructure challenge.

Electric cars are ready for gasoline car use cases, leaving 5-10% or so where EV is not ready. (the places where electric cars ready are not are places where diesel is better than gasoline, though some people do use gasoline anyway). Currently EVs are in the phase of scaling production, but that will take a bit of time.

> We could’ve replaced coal plants two decades ago, however, and reduced emissions by 20% globally

What we would have replaced coal with 20 years ago (and in fact what we did starting more than 20 years ago) is mostly natural gas, which itself pollutes, just less. (depending on what pollution you measure you will get different numbers for how much less) Wind and solar is coming along fast, and is also replacing coal plants (for nearly zero pollution) but we are already scaling those as fast as possible so to claim we could have done more 20 years ago is false.

Electric cars aren’t ready. And absolutely everyone knows that, so why are we bringing it up?

I have been driving an electric car for 4 years, and it serves my needs beautifully. So, no, I strongly disagree. Electric cars are ready for most use cases and many people agree.

how much did you pay for that electric car? They aren’t comparable in price.

It’s weird how we have to have these silly discussions where people must know that they’re obviously wrong before commenting.

Replacing 1.3 billion cars is going to take decades.

The average car lasts about 12 years. Sure some collectors cars will be kept running for decades, but it will take about less than 2 decades to replace essentially all cars if we only made EVs. Only selling EVs is the law set to take effect in a few years.

I'm sure that the law will be rolled back a little when one of the 5-10% usecases where EVs don't work screams loud enough (probably shipping and remote areas), but it will become hard to find fuel for your gas car in a couple decades.

Yes, in the year 2035 the internal combustion engine no longer be sold. Unless we have to push back a few years.

That gets us to around 2050.

27 years from now.

To get to 100% EV usage, yes. In the US, we're already at 5.5% sales of all cars are BEVs, up from a year ago at 3.1%, up from 1% two years ago. In just 5 years a sizable portion of American cars will be BEV, which will create demand for charging and grid usage, which will accelerate the shift. In markets with lots of Tesla adoption, there's already a robust used Tesla market.

I bought it used; the car is ~10 years old, now (still works great and has a much higher resale value than your typical 10 year old car). I have never bought a new car in my life, so I am not really in a position to comment on relative prices of new cars, but they seem quite similar to me, keeping in mind that the electric car manufacturers seem to have been targeting a slightly up-market demographic in general. You also need to consider that electric cars are much cheaper to run than an equivalent petrol car.

It’s weird how we have to have these silly discussions where people must know that they’re obviously wrong before commenting.

I agree . . . oh wait, you think I'm the one who is wrong. How bizarre.

Oh no, don't tell my electric car that's 10 years old

You don't need to use it directly for it to play an important role. If you have the green hydrogen you can convert it to methane or ammonia, which have less issues in storage and transport. Methane is already widely used for bus transportation, although extracting CO2 from the air efficiently enough to create methane from the hydrogen is its own issue.

It's absolutely essential to deal with renewables' intermittency.

Of course this wouldn't be a thing if people wisened up and built nuclear plant instead of wasting it on feel-good project with less than 20% of actual capacity (solar in Germany ... lol).

It's more efficient to cool and liquify air than to generate hydrogen. Store it in a container and let it expand later and you can turn a turbine. There's no issue from an environmental standpoint since you are converting air to air. You can build a cryobattery literally anywhere on earth with off the shelf components.

> It's more efficient to cool and liquify air than to generate hydrogen.

I dispute this claim.

You're assuming that compressing and decompressing air can be done at greater efficiency. That's not that obvious, considering for example that compressing air generates heat that is always lost in the process.

It won't be for long.

Remember, TSMC do chip lithography with a light source that is generated by firing a laser at in flight drips of molten tin metal. The metal releases a wavelength of light that is then able to etc ultra precise lines on wafers for chips. That technology really didn't exist anywhere 40 years ago.

In many ways the plumbing problems of hydrogen are lower bars to get over. And worrying about a lack of infrastructure is also not a big issue. There's lots of engineer's who'll easily transition out of the LNG industry into great jobs working on these problems.

> lithography with a light source that is generated by firing a laser at in flight drips of molten tin

What does EUV lithography have to do with hydrogen’s generation expense, embrittlement of metals, flammability and lack of infrastructure?

Those are all engineering challenges that are broadly constrained by money. I was pointing out that it's pretty common that those types of problems get solved when there's a will. the tide is turning, there's active research into everything you've listed. I'd bet on these problems getting solved.

> engineering challenges that are broadly constrained by money

You’re comparing semiconductor manufacturing, a domain so concentrated one firm (ASML) produces the world’s cutting-edge instrumentation, and so quick-moving our modern paradigms for growth (Moore’s law to venture capital) emerge from it, to building an international piping and shipping system for a novel fuel that speculatively competes with batteries. Nobody is asking if, given infinite time and resources, these problems could be solved. It’s whether the solution would be competitive with what we have. Unfortunately, this blind optimism is baseline for I’ve seen for hydrogen.

Not all problems are solvable. I don't know enough about physics/chemistry to know if the problem is solvable, but I do know there are other unsolvable problems. (faster than light travel is a common one that people think more engineering can solve)

I'm sure people are working on the problem, but the laws of physics limit them.

When something is hyped like fusion for 40 years plus and does not see the light of day, the problems that constrain it, be hard and fruitful.

The problem that constraints fusion is investment capital. You can't expect results while resting below the "fusion never" track: https://i.imgur.com/3vYLQmm.png

No, actually what constrains (DT) fusion is that even if a reactor is produced, no one will want it.

http://orcutt.net/weblog/wp-content/uploads/2015/08/The-Trou...

The chemical reactivity problems of hydrogen are unsolvable in a competitive fashion with other types of energy transport.

The world produces 700 cubic kilometers (at STP) of hydrogen every year. I assure you there are materials compatible with hydrogen, or else this would not be possible.

Hydrogen is not really an engineering problem, its an investment problem. And its simply put a gigantic amount investment for a minimal amount of gain.

There are so many, way, way, way more useful investments you can make. How about doing useful simple things that we are sure can be done successfully and we know for sure will save a lot of CO2 and emission.

Stuff like railway electrification.

If you want to have long range trucks be a thing, electrify the highway with trolley wires.

Hydrogen trains have already shown that they can't compete with electrification traditional or with battery. In trucks they are currently getting their ass kicked by battery trucks. So neither for trucks nor for trains does hydrogen really make much sense.

So why exactly should we invest in a gigantic hydrogen infrastructure. For the maybe 1% market share in trains and trucks?

And its not actually easy to reuse LNG infrastructure.

> And worrying about a lack of infrastructure is also not a big issue.

This is disproved by literally 5000 years of human history where huge infrastructure investments are always a big political problem. And even if projects are clearly of huge benefits they are difficult.

In terms of hydrogen infrastructure, nobody can make a case that it is actually worth it in the first place.

> There's lots of engineer's who'll easily transition out of the LNG industry into great jobs working on these problems.

Or we could make those engineers work on useful stuff like batteries, railway electrification, nuclear, metros and so on.

> Hydrogen is not really an engineering problem, its an investment problem.

The problems stated by the OP are mostly maintenance problems. What is even worse, because you don't just pay once and they go away; you have to keep paying and they will come back if you ever lose focus on handling them.

Railway electrification in the US is a political problem. US freight rail has tiny margins and relies on quantity for profits. Freight rail regularly delays maintenance and upgrades until unavoidable for cost reasons. The US gov't is loathe to make American rail less competitive and force it to electrify, so they continue running polluting trains.

Us freight does not have tiny margin, they pay a large dividends and have spent 100s of millions in stock by back.

> Freight rail regularly delays maintenance and upgrades until unavoidable for cost reasons.

They are gradually running down the infrastructure built 100 years ago into the ground and maintain at the bare minimum while making large profits.

> The US gov't is loathe to make American rail less competitive and force it to electrify, so they continue running polluting trains.

Electrification would make it even more profitable then it already is.

Pretty much every single study done on electrification shows that it pays for itself in a pretty reasonable amount of time. Given that this is something that would still be useful 100 years from now, the US government forcing all first class to invest in this would be of huge benefits both to them and to the US as a whole.

So what's going on here:

Hydrogen Powered Train in Germany: https://www.smithsonianmag.com/smart-news/hydrogen-powered-p...

and here:

Hydrogen Powered Toyota: https://www.toyota.com/mirai/

Malinvestment

We will need many scientific breakthroughs for battery to be viable for non-personal transportation. Then there’s industry, construction, agriculture, etc. These areas of the economy are responsible for about 50% of emissions. We need hydrogen. Not for cars, but for everything else.

"Not for cars, but for everything else." While I agree with you there are some usages where it does make sense and shouldn't be neglected, it is also worth noticing that the current consensus on Metaculus is an estimate of 6% market shares for those usages in 2030: https://www.metaculus.com/questions/552/fcevs-vs-bevs-what-p...

[Edit] consensus established in February 2020

The only thing we need hydrogen for is for chemical industry. And that is enough demand already for hydrogen production plans.

Beyond that you don't need it. Not needed for the waste majority of the things you suggest.

And often there are better solution then using hydrogen directly. For remote construction sites, just turn your hydrogen into anther syn fuel like methanol and run a conventional generator to power electric construction methods.

Hydrogen will be a niche use-case in all the areas you mentioned. And there are so many more important things to do.

For example, for construction, electrified railroads transporting all that gravel around rather then diesel.

Sustainable transportation is not about which fuel is greener. It's about replacing private with public transportation options as much as possible e.g. planning urban areas prioritizing public transit systems and discouraging private transportation

That would be nice, but in the US the people leading public transit don't care about good transit. I know they have limited budgets, but they do too many bad and over prices projects to assume anything else.

Yes as is slowly happening in many urban areas throughout the world. Unfortunately, unlike Europe, the US has spent the last 100 years remaking the American urban fabric for the car. It'll take time and lots of capital to change cultural attitudes in the US to make car driving more acceptable. People continue driving in slowly increasing amounts even in well-planned countries like the Netherlands. In the meantime, pushing existing car uses into lower-emitting forms is key. This isn't an either/or.

Reply on news.ycombinator.com