Hacker News new | comments | ask | show | jobs | submit login

The article looks at the cost of energy over the lifetime of the nuclear power plant. There is no argument that energy generated through fission is very cheap when looked at that way. However that totally ignores the enormous startup costs.

The great thing about wind and solar is that you don't have to build a whole farm. You can start small and keep adding as you come across more capital.

In any case I don't see why one needs to make it a dichotomy. The entities who invest in alternative energy are probably not the same ones who could invest in a nuclear power plant because of the above mentioned startup costs.

It's not clear to me whether fission will come back any time soon but wind and solar will keep gaining in market share.

The startup capital required for nuclear fission is so immense that it doesn't make sense from an economic time-value perspective. The problem with most renewables (excl. thermal and hydro) is that they are incapable of meeting 100% of electricity demands by their nature without better storage technology.

The economics are pushing towards renewables but I feel that nuclear makes more sense for our society in the near-term (we need to get away from coal and other fossil fuels).

99% of the worlds grid storage is pumped hydro because nothing else works out. Pumped storage is 70% to 85% efficient, can come online within 15 seconds, and works at scale.

The reservoir can provide about 13 GW·h of stored gravitational potential energy (convertible to electricity at about 80% efficiency), or about 2% of China's daily electricity consumption. https://en.wikipedia.org/wiki/Tianhuangping_Pumped_Storage_P...

Construction cost: $900 million USD maintenance costs are also minimal.

PS: ~14 GWh for 1 billion ~= 14 MWh for 1 million = ~14 kwh for 1000$. http://www.teslamotors.com/powerwall = 7 kWh for 3k or 2.3kwh per 1,000$ and much shorter lifetime.

The problem is that we have very limited amount of natural reservoirs. Practically every natural reservoir out there already has power plant, is on a desert or belongs to natural reserve.

With tesla solution the problem is going to be limited amount of lithium. There is maybe enough lithium to get a powerwall to every household in U.S. and EU. But rest of the world is fucked.

You can just use two large dams in a row. The Columbia for example has 60 dams on it's watershed some fairly close to each other. https://upload.wikimedia.org/wikipedia/commons/e/ec/Pacific_...

Granted, this would be a significant retrofit, but it's significantly cheaper than starting from scratch. And, assuming the net daily change is ~0 it's not going you don't lose existing power generation capacity or add significant environmental impact.

Also, you don't need very many. China can shift ~2% of it's daily power needs with just one location. Get into the 10-15% range and your done.

You can run the numbers for thermal storage. It's not fantastic. Probably a 60-70% return on input + plus waste heat. However thermal storage are very compact. Rough estimate I get; 100MW for 10 hours with a 2.5 acre footprint. One also gets about 40MW of waste heat which could be used for commercial or industrial heating.

Currently though, complaining about storage has a cart before the horse aspect. Since for the immediate future photo voltaic plants are competing with gas fired peaking plants not nuclear or coal fired base load plants. (If you ever wonder why the Koch brothers really don't like Solar plants it's because solar cuts into the market for natural gas)

Wow, I'm actually surprised that the battery comes that close. When one of the newer battery technology materializes those numbers might reverse.

Vanadium redox flow batteries are shipping, with their manufacturers claiming that they beat Tesla's lithium ion both on cost and geographic density:



Don't forget Li your down capacity after just 3 years and you might get 15 years total. Pumped storage is good for 50 years before a refit and the reservoir is probably good hundreds if not thousands of years.

Of course, pumped storage (or any kind of hydro) is many orders of magnitude more dangerous and environmentally destructive than fission.

The Banqiao Reservoir dam failure alone killed over 170,000 people and made over 11 million homeless.

I worry far more about the hundreds of millions of people living downstream of the Three Gorges Dam than I do about people living near fission plants.

That dam failed when 1 year's worth of rain fell over 24 hours if it had not been there a lot of people would have died anyway.

On net Dams have saved far more than 170,000 lives in china alone. Flood control is more or less a necessity in the modern world adding energy generation on top of that is a minimal risk. ex: From 1998 https://en.wikipedia.org/wiki/1998_China_floods loss of 4150 people, and 180 million people were affected.

PS: Direct deaths where ~26,000 people. The 145,000 died during subsequent epidemics and famine which where blamed on the dam, but that was a convenient excuse and far from the root cause.

I get that it's important to be objective, to look soberly at costs and benefits. But 'only' 26,000 direct deaths? Oh, that's fine then.

Yes, poor word choice. I was comparing 26,000 from a failure in 1975 to 4,150 in 1998 even with lot's of flood control.

Even with lot's of flood control floods still kill some people. But, dams prevent many floods, reduce severity, and generally give significant warning time when there not going to be enough. So, most deaths are from small rivers that feed major ones instead of major rivers overflowing.

Without them, things would be far worse.

Pumped storage probably makes more sense when it is paired with fission than when not. The power companies built this one that isn't quite so scary:


a large amount of startup capital is of regulatory nature. https://www.youtube.com/watch?v=LUdwgbh6he4

I don't think unregulated nuclear is a good idea. Safer reactors would require less regulation and you could argue that even for current reactors our regulation is excessive. The bottom line is that efficient regulation is a hard problem and to use nuclear power we need to either solve that problem or endure inefficient regulation.

It's not that simple. Nuclear regulation in the US has turned (similar to the FDA) into an innovation stifling behemoth... It's a lobbying place and a capital sink and has little to do with safety. If it was about safety, they would shut down current reactor designs yesterday and had started working on Thorium alternatives 20 years ago. The youtube link I posted from TEAC7 conference does explain that quite well.

The current regulation has been pretty effective at what it was intended to do: stop the production of nuclear power plants without the politically unpopular attempt to outright ban them.

People saw Chernobyl and said "none of that in my backyard" and successfully managed to write rules so onerous that they're effectively a ban.

Unfortunately, any plans they had for a solar power revolution in the 70s died when the technology turned out to be outrageously expensive and impractical, and we've been stuck burning coal waiting for the technology to catch up. 40 years of filling the atmosphere with greenhouse gasses because of one spectacular failure halfway around the world and one scare in our own country.

Another irony is the fact that all of our current reactors are old designs and less safe than new ones would be if we were allowed to build them.

Of course all of the political pressure has also killed our waste management plan as well, so everybody has to make due with less safe ad-hoc setups on every site.

I'm not arguing in favor of our current regulation, I'm saying that it is hard to fix and you can't throw it out altogether.

It's not actually that hard to fix. We know how to regulate large industrial operations that use dangerous substances. The problem is that we don't use those regulations, we use a whole different set of regulations that were created by people whose specific intent was to make construction uneconomical.

Example. One of the problems that have occurred is that the regulations change during construction. You spend a billion dollars on construction and then the regulations change and you have to start over. The simple change that the construction rules a plant is evaluated under are the ones in effect when construction began would solve half the problem in itself.

There's a difference between regulated nuclear and over-regulated nuclear. If you take a look at the regulations around nuclear startup/operation you'll see some of the policies are insane.

Compared to fossil fuels, which has been successfully lobbied to be under-regulated, you'll stark differences. If fossil had to even approach the same safety/environmental rigour of nuclear, fossil fuel market share would drop quickly.

Nothing about nuclear fission requires startup capital to be large.

There is a 5-10 year lead time between digging the first hole and opening the plant. That's >$100 million in capital investment for 5-10 years without any return, so the plant has to have a present lifetime value >$140 million the day it opens.

This ignores various risk cases associated with building a plant that drive the return on capital further up.

Source: used to value these types of investments professionally.

You're assuming a very particular kind of nuclear fission machine.

I am unaware of any kind of practical nuclear fission machine that doesn't require a high initial capital investment. Could you link some examples of commercially operating units? It has been a few years since I was in the field but I was unaware of any proven changes to the fundamental economics of the industry.

There's a good chance you can see one at your local university or shipyard. The post to which you're replying doesn't use the terms "proven" or "commercial", so let's not move the goalpost. Then we can talk about YC-funded startup UPower, Fluor-owned corporation NuPower, etc. Studying engineering is also a good way to reduce reliance on examples.

I was unable to find any data on the 'NuPower' technology; could you provide links?

What I found on UPower indicated that it's a nuclear thermal battery, I love that technology but they aren't legal and wont be because of widespread concerns about terrorism and radioactive contamination.

Also, my educational background is in engineering and I have worked on determining whether it's financially feasible to build power plants for a living. I would really love it if you could provide some evidence for your arguments.

NuScale Power, sorry (mobile typo). UPower are making a reactor but some sources have incorrectly called it a battery. There are no laws in the U.S. (or any nation I'm aware of) that make small reactors or nuclear batteries illegal. My only argument so far is "Nothing about nuclear fission requires startup capital to be large", which is trivially true.

The design and safe construction does.

Very informative read and eye opener on this topic: http://thorconpower.com/costing/should-cost-versus-did-cost

Additionally, many people are completely unaware of the vast variety of fission machines of all shapes and sizes -- rocket engines, ramjet engines, space power reactors, research reactors, isotope production reactors, submarine power reactors, Army mobile power reactors, etc. -- that were produced and operated 50 years ago. To say nothing of the variety we could build with today's metallurgy, CAD, etc.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact