Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

UPower (YC S14) Is Building Nuclear Batteries

techcrunch.com

Post reply on HN
Print view View original

41–50 of 123 posts

Re: UPower (YC S14) Is Building Nuclear Batteries

#41

Post by donttrustatoms »

I am one of the UPower founders. The key takeaways here are that the main hindrances to novel nuclear are (1) meeting regulatory and safety necessities without huge testing costs, which also require a customer interested in purchasing, and (2) competing with the grid, (3) huge scientific/technical risks if not yet proven.

UPower meets these by (1) being small enough to build a very tough, full scale emulator for testing, never before done in nuclear. (2) not competing on the grid, which gives immediate, in need customers which is also necessary for regulatory process to complete in (1). (3) using technologies and materials with decades of experience, in a new way, eliminating the huge uncertainties of many startups you see in the news.

Re: UPower (YC S14) Is Building Nuclear Batteries

#42

Post by XorNot »

ekianjo wrote: post #19
damian2000 wrote: post #14

Earlier quoted context omitted.

Except you just have to look at Fukushima to see that things can still go badly wrong, resulting in a meltdown of 3 reactors. And this happened in arguably one of the most tightly regulated and technologically advanced countries in the world.

I live in Japan and I don't consider that the outcome was as catastrophic as the media said it was (they are always ready to jump on the apocalyptic trigger). Everything went wrong (even the backup generators as they were blown away by the tsunami) and it was still reasonably contained and the levels of ambient contamination are nowhere as high as they initially predicted. EDIT: > most tightly regulated and technolog…

Also if there's any lesson here, it's that government intransigence is the real threat here. They don't want to build new reactors, they don't want to shut down old ones, but they still need electricity too.

And of course consider the scale - something like 8,000 people died from that tsunami. Oil refineries burned for hours (how much heavy metal got dumped into the ocean and soil?)...how many people did Fukushima actually kill?

Re: UPower (YC S14) Is Building Nuclear Batteries

#43

Post by beefman »

rjdagost wrote: post #39

There is a company called Gen4 Energy (formerly Hyperion Power Generation) that has been pursuing the "nuclear battery" concept for 7 or so years: http://www.gen4energy.com Regulatory approval has been elusive, and that's for a team of 30+ year nuclear industry veterans. The challenges in this industry are brutal, I hope UPower has deep enough pockets and extreme patience to shake up the industry.

They started with a good-sized design, but have scaled it up a couple times since then. They're no longer in mass-production territory.

Re: UPower (YC S14) Is Building Nuclear Batteries

#44

Post by ekianjo »

XorNot wrote: post #42
ekianjo wrote: post #19

Earlier quoted context omitted.

I live in Japan and I don't consider that the outcome was as catastrophic as the media said it was (they are always ready to jump on the apocalyptic trigger). Everything went wrong (even the backup generators as they were blown away by the tsunami) and it was still reasonably contained and the levels of ambient contamination are nowhere as high as they initially predicted. EDIT: > most tightly regulated and technolog…

Also if there's any lesson here, it's that government intransigence is the real threat here. They don't want to build new reactors, they don't want to shut down old ones, but they still need electricity too. And of course consider the scale - something like 8,000 people died from that tsunami. Oil refineries burned for hours (how much heavy metal got dumped into the ocean and soil?)...how many people did Fukushima ac…

> how many people did Fukushima actually kill?

Hey, we already know the answer. No-one. Not even the guys who went and exposed themselves to high doses of radiation to keep the reactor under control. And people who may have been contaminated are being strictly followed for potential leukemia and other cancers. Following the events, I'd say the risks regarding the nuclear power were relatively well managed.

Re: UPower (YC S14) Is Building Nuclear Batteries

#45

Post by donttrustatoms »

omegaworks wrote: post #29

>with some processing Weasel words. Reprocessing spent nuclear fuel introduces a MASSIVE regulatory boondoggle.

Great, great call to point out.

The author very unfortunately did not understand that there is NO "reprocessing" necessary for our reactors to reuse fuel. Why, oh why, don't we get to preview what they write. Wording is so important in such a massively misunderstood space.

It is a proven process utilized previously in fuel manufacture, and simply adds a little new fuel to existing used fuel. It does not involving isolating any portions of the fuel as done in reprocessing (which is the main "proliferation hazard" of reprocessing).

-COO UPower

Re: UPower (YC S14) Is Building Nuclear Batteries

#46

Post by msandford »

jddw wrote: post #27
tim333 wrote: post #7

Hmm. 2MW/reactor means you'd need 500 of the things knocking about the country to equal one regular power station. Not that great for keeping things secure.

Right, the goal is to start very small, and off grid. Then we scale and grow to grid size.

And there's nothing that would preclude you guys from simulating geothermal with them.

1. Dig down 50-100 feet

2. Make a foundation and catch pan

3. Put in an elevated platform to hold containers

4. Lower 5-50 containers onto elevated platforms

5. Install first order heat exchangers downstairs

6. Put "roof" on

7. Cover with 30 (or more) feet of dirt or whatever you prefer

Perhaps install a bridge crane and a shaft to the surface so that you can swap power units out without digging down again. Just make incredibly massive covers (like 10ft thick concrete) without cranes on-site at the surface so that terrorists wouldn't be able to hijack stuff once they made it through the front gate.

It might look something like this: http://www.theregister.co.uk/2009/04/10/google_data_center_v...

The idea of simulating geothermal is that you'll have a bunch of power generation equipment on the surface that runs on heat. It just happens that it's not a "natural" geothermal well but a man-made one.

Re: UPower (YC S14) Is Building Nuclear Batteries

#47

Post by donttrustatoms »

tim333 wrote: post #7

Hmm. 2MW/reactor means you'd need 500 of the things knocking about the country to equal one regular power station. Not that great for keeping things secure.

To elaborate, The idea is that this is our Tesla Roadster to get to disrupt an out of touch industry, and to get to in-pain, in-need markets, and to get through regulation quickly. The next step is the scaled, grid-competitive reactor- our "model E".

The first ever commercial reactor was a micro reactor, just 5 MWe, outside San Francisco started by GE. These huge companies have lost their way in innovation and off-grid markets have changed dramatically in recent years.

How appropriate is it then for a new Silicon Valley company to start again, disrupting with a new competitive micro reactor...!

Re: UPower (YC S14) Is Building Nuclear Batteries

#48

Post by beefman »

pjscott wrote: post #31
beefman wrote: post #16

Earlier quoted context omitted.

It's a reactor, not an RTG. One reason I don't like the term "nuclear battery". The term is also annoying here because it's clearly an attempt at spin. The cultural situation with nuclear is beyond spin, and requires direct confrontation.

The term "nuclear battery" here refers to the operational model: the modules are shipped out from a factory, hooked up on-site, run with minimal intervention, and then shipped back after several years for refurbishing and refueling. This centralizes construction and maintenance, which is important when the reactors themselves are so small.

I know. I'm arguing against this use of terminology. Nuclear batteries are things, and nuclear reactors are things. Describing your (excellent) operational model should be straightforward and should not require calling a reactor a battery. What you are really doing is demurring from confronting misunderstandings about nuclear energy.

Re: UPower (YC S14) Is Building Nuclear Batteries

#49

Post by idlewords »

donttrustatoms wrote: post #41

I am one of the UPower founders. The key takeaways here are that the main hindrances to novel nuclear are (1) meeting regulatory and safety necessities without huge testing costs, which also require a customer interested in purchasing, and (2) competing with the grid, (3) huge scientific/technical risks if not yet proven. UPower meets these by (1) being small enough to build a very tough, full scale emulator for test…

I think the key-est takeaway is the need to persuade people not to be terrified of nuclear energy. That's likely to require an effort on a par with any technical advances you make.

Re: UPower (YC S14) Is Building Nuclear Batteries

#50

Post by idlewords »

ekianjo wrote: post #10
chrissnell wrote: post #2

Are size and efficiency the most pressing issues to solve in nuclear power? I don't know a thing about nuclear issues but from the layperson's perspective, it seems like security is the biggest problem that keeps nuclear power from being used everywhere. These containers are essentially dirty bombs without detonators and must be protected as such. This might work for the military, where you could secure one on a rest…

> essentially dirty bombs without detonators and must be protected as such. No, it's completely wrong. That's not how a nuclear plant works. And as you can see, a nuclear plant does not "explode" when it goes amok. The chain reaction is very tightly controlled and there are multiples redundancies in case things don't work as expected.

Nuclear plants certainly do "explode" when things when wrong. I assume you used scare quotes because you were thinking of a nuclear detonation, but there was plenty of damage done by the non-nuclear explosions at Chernobyl and Fukushima, to cite the most famous examples.
Post reply on HN