Tiny nuclear battery be commercialised? | Hacker News Reader

Tiny nuclear battery be commercialised?

slguardian.org

28 pointsrakkhi58 comments

Nickel-63. Not too bad. It's a beta emitter. The risks are low, unless you eat the thing.[1]

"By placing a 2 micrometer-thick nickel-63 thin film between two diamond semiconductor converters, the decay energy of the radioactive source can be converted into electrical current, creating an independent modular unit." Nice. Packaged that way, it's not likely to become airborne pollution. Even in a fire.

As is usual for battery articles, the claims of applications are excessive. This could be useful for many small items that need battery replacement, but is unlikely to be enough for drones, etc. Watch-sized wearables, maybe.

A similar concept was announced in Russia in 2018.[2] The one from China seems further along.

[1] https://ehs.princeton.edu/laboratory-research/radiation-safe...

[2] https://phys.org/news/2018-06-prototype-nuclear-battery-powe...

> many small items that need battery replacement,

environmental sensors maybe?

I have real doubts about this tech being viable on the market. For one thing there's a huge psychological barrier for many people, as anyone who has worked in nuclear power can tell you. There's also the issue that this relies on Nickel-63, and while unfortunately info on total demand/production is significantly paywalled, I don't think it's enough to support something like this. Outside of chromatography and explosives detection, there isn't a huge market for it, which implies low production.

It is also true that it isn't as dangerous as many other isotopes, but unless these batteries are sealed to last 100+ years, disposal/recycling is going to be an issue. That will add cost, even if the production of Nickel-63 can be scaled up from (probably) 10's of kg's to 1000's of kg's annually.

> Unless you eat the thing

Normal batteries aren't good for you if you do that, either.

They are mostly sealed though, so you'd probably just pass it?

Not for the button batteries. They are especially dangerous, because the physical distance between the poles is small, and your digestive tract is filled with electrolytic solutions. They can cause chemical burns and even death.

https://www.healthychildren.org/English/safety-prevention/at...

The inside of your body is wet, so even low voltages can make a complete circuit. Something as simple as a CR2032 button battery can burn you from the inside.

https://www.youtube.com/shorts/hPKSJD3xbok

https://youtu.be/qQcxNTNRME4

https://www.poison.org/battery/stats

There are some number of deaths a year from this.

Weirdly I was listening to a radio show about the Voyager space craft a few days ago and they talked about the nuclear batteries powering those, making me wonder if this was something that could be used here on earth. I know absolutely nothing about this tech so thanks for this useful tempering of the hype.

This tech has been used on Earth from the 50s. Notably it has already been used in the 70s in the heart pacemaker batteries implanted right into human body! Any hype is unwarranted. https://en.wikipedia.org/wiki/Atomic_battery

Ah ni-63.

It’s a pure low energy beta emitter. Can’t detect with a Geiger counter, best detected with liquid scintillation swabs. Makes it annoying to detect.

As long as it’s outside the body it’s harmless. Ingesting or inhaling is not recommended, will harm you once inside.

Also as a side note, battery is a bit of a misnomer for nuclear batteries, they’re more of a constant current source. Use it or lose it.

Given the structure of the thing, it seems closer to the original voltaic pile

The decay doesn’t stop when the circuit is open

How is this the “world’s first micro-atomic energy battery”?

Didn’t the USA do this in 1966?

https://en.m.wikipedia.org/wiki/Plutonium-238#Nuclear_powere...

That thing used the heat. A Beta battery just has new electrons appear, "magically".

https://en.wikipedia.org/wiki/Betavoltaic_device

Thanks for the article. But then it points to the Betacel, USA 1970s (60 pacemakers implanted by 1973).

https://en.m.wikipedia.org/wiki/Betacel

I’m not trying to dismiss a wonderful invention, I just want journalistic integrity. It’s not a first if someone already made a product in 1973.

100 uW is not a lot. You'd have to collect it and burst to run anything more then a couple of gates.

It's enough to keep a low power microcontroller and maybe a low power sensor or two alive. You can get chips and sensors that will work with that budget, you just have to be very conservative and sleep aggressively. E.g. there are PIC chips that consume 25 uA/MHz at a couple volts while active (not sleep) and can do useful tasks: https://www.microchip.com/en-us/solutions/low-power

You could also use more than one of these sources.

Couple of problems here. So let’s say you use one of the chips running at 25 uA (lowest run mode). First of all the does not include any peripherals. UART/i2c/timer/adc/dac don’t run in that example. And god forbid you want to do RF! Then we also need to allocate power to rest of the board. The power supply is going to have quiescent (30uA if not more… probably more as it’s probably a HV source). Then you need to allocate power to sensors and rest of board, and the uA go fast.

Yes that's what I mean that you have to take great care, I was trying to say that it's not easy to engineer around these, just feasible. You still want some capacitor buffer, work in short bursts, sleep everything aggressively, etc. And while there are definitely useful applications they aren't gonna be as revolutionary as the headlines that always come up about these things seem to imply. But you can do more than power a few gates.

Regarding power supply, I'm not sure what these exact ones are, but other betavoltaic batteries are not high voltage. They're in the range of a volt or two and these are probably the same since these work more or less the same way as every other betavoltaic battery with a bit of extra proprietary sauce.

Edit: Also regarding RF, you can do passive RF via backscatter or a complementary RF energy harvester. Imagine a temperature sensor that logs data a few times a day powered off the battery, and someone with a handheld scanner comes by and reads it out every so often. Or an active tamper seal that spends most of its life making sure a wire isn't broken. Just brainstorming, but something like an RFID tag that can do stuff while it's not being interrogated is the main proposed use I've seen for these.

I mean, it's about 100B of them to power a high end Tesla at peak output. You could likely just use 1 billion of them and couple them to a capacitor bank.

That's not quite as outlandish as it sounds, given that they're very small, 15x15x5mm

And of course, this is an early gen product, not ridiculous to assume they'd be able to make these smaller, and with higher output, in the future, so we very well might get there in time. In terms of applications, there are a ton of viable applications between "occasionally used low power environmental sensor" and "car" of course.

1 billion would fit into a cube of 1000 units long, wide, and tall. This would be 15 meters by 15 meters by 5 meters, or 49 feet by 49 feet by 16 feet, or much much larger then the entire car.

Workable submarine though! Could a submarine get by with only 400 horsepower?

Likely, submarines are slow and quiet, and this would be very, very quiet given there wouldn't be pump noise to worry about. That's the big hurdle with the current subs, you have to circulate water in and out of the reactors to keep them cool, and for an attack sub, you don't have enough insulation/mass/space to keep them 100% quiet. The missile subs are much quieter, though, once you cook off a missile...well that's a very loud thing to do in a very quiet place.

There is still the problem of the rest of the sub's power needs. I can't imagine they're terribly efficient given they have a reactor available...

Yes, but, say you can get 1 watt per module, or 10, suddenly it starts to become more reasonable for many things as long as the modules stay small.

Sorry to be critical/cynical, but isn't this just a resurfacing of the widely-discredited nuclear diamond battery? [1]

This looks identical. 100µW, diamond etc.

How they plan to go from 100µW to 1W is dubious at best. They'll stack 10k of these? Ok. Maybe.

Then how they go from that to powering drones? Come on. ~100W minimum for a small consumer drone with a camera. So stack 1 million?

[1] EEVBlog: https://www.youtube.com/watch?v=5M5MF6KE-jY

I wonder how prohibitively expensive (or not) this will be. The other comments talking about lack of applications they're able to think of on the spot are rather besides the point, the only questions are whether this will be produced at scale and at what cost.

Title grammar nit: maybe change "battery be" to "battery to be"?

Yarrr, it be fine.

From the article it seems to have only advantages. Is there any downside? Why shouldn't it be used in every portable electronics?

Running the math suggests 0.1 mW. Which... the specs of many uCs I play with can theoretically drop down to 100uA or less of current (especially if I have capacitors to store energy while the uC sleeps).

But ugggh. This won't be easy to design around. A lot of capacitors will give you 10uA of leakage already. Assuming 3V uCs, that'd be 30uW of precious energy just in leakage current.

I think most people would rather use a AA battery. Is it really that big of a deal to replace the AA batteries every 10 years?

These things are awful from both a size-to-power and weight-to-power ratio.

In other words, any application where size or weight is an issue (nearly all consumer portable electronics such as laptops, phones, tablets, power tools, and especially cars), these won't be good enough.

The last time I heard of a radioactive battery (btw, this concept is not new), I'm pretty sure someone did the math and found that you'd need one the size of a typical car battery just to get enough wattage to power a phone. To power a car, it'd have to be the size of a semi trailer.

That said, their life is certainly huge. If size and weight isn't an issue, they can be an incredible boon, but I imagine at the scales I'm thinking of (like needing a few kilowatts to power a 5-person outpost somewhere), the cost ends up being the major factor.

We can safely assume that anything containing radioactive material that is not tightly controlled will not be commercialized.

ionization smoke detectors and radium clock dials come to mind, neither of which are banned in the U.S., although you won't find any new radium dials.

> ionization smoke detectors

It would be unpatriotic to ban americium.

They are not banned because it's very old tech from the times where the regulation was much more lax.

Closer analogy would be tritium vials/dials, which are also beta emitters, and are also not banned in the US.

Aren't batteries going bad the main reason people replace devices right now? Sealed in so that it's difficult to replace them or get them replaced?

I feel like that's why we wouldn't see this. It's like a cure for a disease that people are making a ton of money treating instead of curing.

while I don't think these are going to work unless they can get a 5 watt battery the same size as a phone battery, if they actually have a 50 year lifespan like the company claims that would be WAY more than the lifespan of the device. If it actually existed you'd actually have the opposite problem. "My battery is fine but this device just isn't keeping up, can I transfer my battery to my new device?"

In your mind, someone makes a battery that works, then sells it, then what happens?

People don't buy another one for a long time and they go out of business? It's not like planned obsolescence isn't built into so much of what we own and buy today. I guess maybe I don't understand the question?

I guess maybe I don't understand the question?

Correct, I'm glad you asked. Most batteries are embedded in electronics and have dozens of different form factors. Are you under the impression people buy a single battery and replace it when it wears out?

Under this logic wouldn't rechargeable battery makers have put themselves out of business a long time ago?

Do you not understand that rechargeable batteries have a lifespan? A finite amount of cycles where they're viable to recharge and continue use? If you have an iphone, have you checked the reported battery health on it?

The point is, if something is producing power and never needs replaced, it never goes bad. As in, it's not just rechargeable it's effectively infinite from the perspective of a human's lifespan.

So, no, rechargeable battery manufacturers aren't out of business because after a couple hundred recharge cycles they need replaced still.

Similarly, with devices like phones where the batteries are usually sealed in, the batteries go bad after a year or two and frequently you're met with "Should I pay $150 to get the battery replaced or $1000 for a new phone?" and, after 2 years sometimes it's more practical to replace the phone. But you wouldn't be met with that if your battery lasted "forever" you'd replace the phone when you couldn't stand the age anymore or when it fell out of support or when you broke the screen. Most of those things take way longer than the 2 years or so of daily recharges that it seems phone batteries last.

If you can't understand why, think about devices that are plugged in. (which is effectively what this would be since it generates power) I've had the same computer running nearly 24/7 for about 6 years. In that time, I've replaced some batteries and bought new phones/tablets/etc due to batteries that started dying at 20% or not fully charging.

You realize this would have nothing to do with phones right? It would be for long term low energy devices. That's like saying the solar strip on a calculator would put calculator makers out of business so Big Calculator won't let it happen.

How many 5 volt amps hours is 8.64 joules per day?

That's enough to lift 8.64 apples 1 meter high once a day.

for those confused, 1 meter is approximately 0.66 Ariana Grandes

https://twitter.com/GatorsDaily/status/1524768862465576962?l...

Ampere-hour is a unit of electric charge, not energy

5V amp-hours is watt-hours, which is joules, which is energy.

you are right, somehow I skipped the "5V" part. My bad.

The article says it produces 100 microwatts continuously for 50 years, with a 1 watt version coming in 2025. Just make a stack of 5 of those.

5000 of those, right?

8.64 joules = 2.4 mW hours. Dividing by 5v gives 0.48 mA*hrs, so you can run a 5v load for 1 hr if it takes 0.48 mA.

Take the easy route when it's presented like that! 2.4 mW hours is 0.1 mW days, dividing by 1 day (i.e. doing nothing / applying the "per day" part of the original units) and 5v gives you 0.02 mA.

You could also just go back to the sentence just before "8.64 joules of energy per day" and see that it produces 100 microwatts (or 0.1mW).

What happens when the battery explodes?

I don't think it's even possible. The term "battery" may not be the apprpriate one to use here, it's more of a "constant electrons generator", than a battery per se. Betavoltaic batteries just keep "creating" electrons, at a constant rate, for a long time. You can't simply "create" 1000 times more electrons on a single moment and release it suddenly. These things are a lot safer than lithium ion batteries when you look it from this angle.

Reply on news.ycombinator.com