Comment by epistasis | Hacker News Reader

Comment by epistasis | Hacker News Reader

Parent Full thread

Compare your best estimates of land usage to the amount of land the US uses to grow corn for ethanol.

Or compare it to the amount of land used for parking. Or for roads. Or for roofs of buildings.

I've never, not once, ever seen a comparison of this sort that sounded the least bit scary.

2. I'm a bit flabbergasted that you think we can scar nuclear production to what we need. We can't even scale nuclear production to hit replacement rate of our existing reactors. Nuclear does not scale, because it can not be built.

And I'm similarly flabbergasted that you think storage would hit scaling problems, when we build new factories for storage all the time, use new chemistries all the time to reduce the use of limited input components.

Storage production is growing at a massive pace, problems are getting solved all the time. Nuclear is shrinking all the time, gets more expensive all the time, and we can't even figure out basic concrete pours and welding. These industries are hugely contrasting, but I don't know how you can reach your conclusions when looking at the reality of storage and nuclear as they are in the real word.

Nuclear is actually intentionally crippled through legislation under over-hyped concerns over dirty bombs accidents (for the right and left politically inclined respectively).

Nuclear can be done safe, efficient, cheap and at-scale. All it requires is to allow for the secondary market of spent fuel like was originally started when nuclear first starting to get rolled out. The security concerns can be handled in other ways than crippling nuclear from being self-sustaining I think.

Can you explain this a bit more? I don't see how these things could have helped any of the construction failures that have plagued all nuclear attempts.

Operating costs are low for nuclear, it's the construction that is the problem.

Pay money to Rosatom, they'll build nuclear plants for you. If US lost its nuclear competency, it does not mean that everyone else lost.

Correct answer. Floating nuclear power plant recently went operational , supplying arctic city with electricity ; it was built in two years, top to bottom. I wonder if couple of those in the Gulf of Mexico could've secured oil refineries' energy supply, allowing grid operators to divert more conventionally generated power to residential areas, diminishing blackouts

The floating power plants from Rosatom are like 70 megawatt, a pittance vs the blackouts Texas saw. The root problem in Texas was insufficient winterization of the grid, caused by the incentives of their energy market approach. This would have still been the case even if Russia had 100 reactors floating off the Texas coast ready to hook up.

There's significant latency in oil supply chains. Weeks if not months. Keeping a refinery online does not somehow keep a power plant online or the lights on in the cities in the proximate sense. Oil refineries are irrelevant to what happened.

... Not to mention the decommissioning costs

I wouldn't be so sure about cobstruction costs either. Reactors are being built elsewhere in the world all the time, and I doubt foreign finance guys are less smart or safety codes are less strict

I used to be of the same opinion, but I gotta say that Fukushima really did have an impact on me (on the long-run, it didn't have an immediate impact). But I have been wondering, if an incredibly advanced nation such as Japan faces very serious issues with their nuclear power-plants, is it really the right choice? Can we really guarantee that nothing goes seriously wrong, ever? Cause that's the standard those plants have to operate under, what if there's an earthquake in Indonesia right next to one of those nuclear power plants, is Indonesia in a better position to deal with such a disaster than Japan?

Nuclear power plants have zero tolerance for serious accidents. And zero-tolerance is never a good margin to operate on.

It's worth noting that, even taking second-order effects into account, the total number of deaths from nuclear accidents is actually quite low compared to other methods of generating electricity. https://www.youtube.com/watch?v=Jzfpyo-q-RM

Nuclear seems more dangerous for the same reason planes seem more dangerous than cars—when something goes wrong, it's a big news event.

Data is cherry-picked. For example, if 60 000 people die because of Chornobyl radiation to 2005 (estimation from video) doesn't mean that they will stop dying after 2005. How many people will die after 2005? Video gives no answer.

Also, the video indirectly suggests that we can replace flood protection with nuclear energy. :-/ Dams are built mainly to protect population from floods, so they SAVE millions of lives every year, while also generating electrical energy, which is used to upkeep dams.

Banqiao incident is pictured like Chornobyl`, but, in reality, a bunch of people lives was SAVED by lower dam. The flood broke 62 dams. Just imagine what may happen when 10 nuclear reactors and/or nuclear waste sites will be flooded instead.

The main problem with nuclear is that when something goes wrong, it last for way longuer than any other form of catastrof.

One of the other comments was talking about the land-use impact of renewables. Nuclear exclusion zones are pretty bad from this perspective.

Aside from Chernobyl (which happened some 35 years ago), nuclear exclusion zones are pretty small & don't have a significant impact (Deepwater Horizon well was arguably larger in terms of the area impacted).

Fukushima happened after the biggest Earthquake in over 1000 years of recorded Japanese history hit at the right spot to produce an enormous flood wave at the reactor.

I'd be careful to draw conclusion from such an outlier.

And radiations in Fukushima will last for the next 1000 years. Do we have the same understanding of "zero error tolerance" ?

At the pace we're going, we're tolerating to make the whole planet a burning hell for the comming millions of years. So I would say we have a pretty high tolerance to disaster, as long as it's happening to our children rather than us.

Considering the magnitude of the damage as well as how rare it is, is a very valid point.

And it's sometimes forgotten that 10s of thousands of people had to leave their homes.

Though some googling reveals that the main contaminant is Cesium-137, which has a 30 year half life. That will fade into background radiation long before that, even ignoring cleanup operations.

They've already reopened parts of the closed area.

> outside the geographical areas most affected by radiation, even in locations within Fukushima prefecture, the predicted risks remain low and no observable increases in cancer above natural variation in baseline rates are anticipated.[13] In comparison, after the Chernobyl accident, only 0.1% of the 110,000 cleanup workers surveyed have so far developed leukemia, although not all cases resulted from the accident.[14][15][16] However, 167 Fukushima plant workers received radiation doses that slightly elevate their risk of developing cancer.

The hysteria of your statement doesn’t appear to be substantiated by a dispassionate look at the facts.

Similarly, I’d be careful about discounting it simply because it was an outlier

It's been years so I can't recall the name or find a source, but I recall reading at the time that there was in fact an engineer working on Fukushima's design that loudly protested that the sea wall and related preparations were inadequate based on historical records. Basically that they planned for the 100 year event but should have planned for the 1000 year event. It also wouldn't have been a huge cost burden. Putting the gensets on piers would have prevented the worst of it.

This is especially worth considering in light of what recently happened in Texas. Say we get a really good setup: teams of nuclear and process engineers doing the best work of their careers, money is no object when it comes to safety, etc. and build infrastructure safe against once-in-10ky events and everything is executed perfectly.

All of that costs money, and sooner or later some MBA / politician is going to ask why we’re spending so much on something which hasn’t come anywhere near a major event – surely we’d be just as safe spending 10% less and using the money for something else like rewarding shareholders? Setting up governance structures which can resist that kind of pressure for millennia isn’t something we have much experience with.

Actually Texas makes a great argument for more, smaller and distributed reactors.

Micro molten salt reactors make a lot of sense if looked at rationally. Can burn what we currently call "waste" too.

That’s a different question. The real underlying issue in Texas was that every producer had an incentive to be cheaper but not to winterize, because they’d immediately see wins from being cheaper on the marketplace and could gamble on going years without a winter failure — by which time the person who made that call might not even be working there any more.

When talking on this kind of scale about a system level failures, the social issues are going to dominate unless it’s something like a fire-and-forget space probe. If you want to see nuclear succeed, the first thing I’d be thinking about is how to ensure that someone doesn’t cut safety to boost their profit margin, accounting for possible future regulatory capture and austerity budgets. Since nuclear is never even going to be within one order of magnitude as distributed as wind/solar, I’d especially think about what would prevent some unscrupulous CEO from pocketing the difference comfortable in the knowledge that their family doesn’t live anywhere nearby.

Given the risk of natural disasters, which place seem safest? Downstream of a dam for a hydro plant, or near a nuclear plant?

188,000 people were evacuated in California following some unexpected rainy weather.

Nuclear loses to alternatives on a pure ROI basis. Even in places where political opposition or legislation is a total non issue (eg China) you see exactly the same bad math when it comes to the $.

I feel like a broken record posting this in every thread about nuclear. There's too many people on this forum that have some sort of smug attitude about not being "scared" of nuclear power and don't bother to learn what's actually constraining things.

I feel like a broken record when people always focus on ROI and up front costs and completely ignore the millon+ people that did every year from fossil fuel usage

Thank you for making this point. I completely agree. I make the ROI argument here because it's the one people listen to, sadly. Living next to a coal or oil fired power plant is one of the worst decisions you could make health wise, particularly for children.

Nobody was under any impression that decarbonizing the energy sector would cost more than continuing to emit carbon.

The reality is that renewables have no plan to eliminate carbon emissions that isn't contingent on a massive breakthrough in energy storage. The US currently had under 10 minutes of energy storage. Renewables offer better carbon emissions reductions on a dollar by dollar basis until they saturate the market during peak production. Then they don't help.

Sure, nuclear is more expensive but it is the only existing solution that aims to actually eliminate carbon emissions. Not just supplement fossil fuels.

Go through Lazard's levelized cost of energy storage slides. We're currently hitting the tipping point where renewables + storage wins based on pure $ without subsidy.

It's completely disingenuous to claim that nuclear is the only solution that aims to eliminate CO2 emissions.

And what form of storage are you using? There's only 20 minutes of storage available via lithium ion batteries with current mining methods. Others like hydroelectric are geographically dependent, and are large concrete and earth moving projects much like nuclear plants.

Lazard uses current commercial cost figures. If people actually tri d to deploy grid storage on a wide scale, prices would skyrocket as supply fails to match demand.

The sweet spot in most models is a mix of hydro, batteries, and H2.

While scaling up the industry isn't trivial, it's absurd to claim it's impossible or that demand outpacing supply will somehow reduce investment in greater supply.

Well let's hear it then. What is the mix? How many GWh of each storage type are we going to use? Typically plans for renewables + storage assumes some magical silver bullet like hydrogen or thermal storage that will make storage nearly free. Except neither of those have actually seen commercial deployment so they're no telling if they'll actually pan out.

You haven't actually given a plan to provision this much storage - or even specified how much storage we'll need for that matter - so yeah I'm still very comfortable in claiming it is not possible. At least not without some massive breakthrough in storage technology, which doesn't yet exist.

Hydrogen could achieve something like 84.8 PWh: https://www.pv-magazine.com/2020/06/16/hydrogen-storage-in-s...

While it's unlikely to reach this value, it's such a high figure that actually can solve the energy storage problem altogether.

If someone actually finds a way to make a feasible storage solution from it. Nobody is installing hydrogen energy storage right now.

Again, the answer is point to an unproven form of storage and assume it'll be a silver bullet.

It's underground salt caverns. This is extremely feasible and has been done for decades with natural gas.

How many Gigawatt hours of underground H2 storage are currently deployed? What company can I call to install 1 GWh of hydrogen storage?

Ah, the good old "if it hasn't happened yet it can't happen" argument.

I'll come back at you: what technological component involved in underground storage of hydrogen is not already demonstrated?

It's been done for decades: https://en.wikipedia.org/wiki/Underground_hydrogen_storage

It's pretty well-established tech at this point.

It's not that nuclear cannot be built. It's that America is incapable of building... Pretty much anything in the 21st century. If the interstate system were to disappear tomorrow, for example, we simply wouldn't be able to rebuild it.

It's not because concrete is some arcane technology that isn't cost effective, or doesn't work. It's because we just aren't very good at building things anymore.

That and we can’t displace black people anymore to build national highways thru residential areas. /s

However I do agree. Our politicians prefer to wank over unimportant things instead of working on things that actually would benefit the people. And the government contracting system is horribly broken.

There is no incentive to putting in the hard work to accomplish projects like before.

This isn't a great argument for building more nuclear plants. It might be an argument for hiring someone else to build them...

New interstates are being built all the time. They are under construction today.

Yes, and they are built at absolutely ruinous costs.

Reply on news.ycombinator.com