Practical power beaming gets real | Hacker News Reader

Practical power beaming gets real

spectrum.ieee.org

89 pointsgmays65 comments

The article jumps between multiple experiments (some laser some microwave), but if you're interested in the the recent 10 GHz microwave experiment the authors were a part of (and which was on the HN frontpage recently) -- 1.6kW delivered over 1km -- the details were published in [1].

Very relevant datapoint: the transmitter source was 100kW, fed into a 5.4m dish. The receiver array was 2x2m. The end to end efficiency, which this article pointedly avoids, is on the order of 1-2%, over 1km.

[1] https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=966...

The laser system shown in the top photo is similar in efficiency. And, trust me, the engineers are very aware of it. It's also frighteningly expensive, with a large number of exotic things.

(Source: I work with the guy who designed a big chunk of the receiver circuitry, among other things.)

100kW turning into 1.6kW is just gross, go buy ten 380W STC rated 72 cell monocrystalline Si photovoltaic panels, a really good charge controller and 10kWh of good quality battery storage instead.

Forget about about wireless power transfer efficiency, the normal wireless power amplifier (PA) efficiency for cellular wireless applications (or any wireless systems) is probably just around 10 - 20%. This inefficiency is terrible not only because the use of modern linear wireless power amplifier but also due to other factors.

For starter, there is a significant distance of wired connection between PA and the antenna that resides on the tower (yes wireless PA still need wires to function). This distance create delay effect on the signal and suddenly this unaccounted impedance mismatched (because delay is phase), and changing phase results in totally different impedance on the Smith Chart even for exactly the same signal amplitude.

Secondly, wideband signal impedance matching (tens to few hundreds of MHz that are now the norm rather the exception) is impossible to do in analog domain as practiced by the industry and putting anything and everything not working as the mythical "memory effect".

Can't wait until the climate change people get into the matter and realize that about half of the wireless systems power budget is due to the inefficiency of the freaking PA and all hell breaks loose.

One of the big problems with the current power grid is transmission loss, resistance in electrical wires over long distances eat ~5% of the generated power. This makes it infeasible for example to have large solar installations in the American southwest transmit power to the Northeast.

Does anyone know about theoretical minimum losses from this kind of power transmission?

a current induction charger for a phone, that operates over about a millimeter of distance, is at best 70% as efficient as charging with a wire.

The fact that this article omitted all details about efficiency, and omitted any numbers that we could use to calculate it, say it's probably not beating the standard induction charger's efficiency.

I think this technology is more in the stage of "It's possible to do this!" rather than "it's practical to do this"

Word games are common in advertising and that has spread to journalism as well.

“it’s impractical to shop for a house in a war zone” and “it’s impractical to buy a house in San Francisco” are two very different definitions of “practical”.

It feels like that’s what’s going on here.

Very roughly minimum loss is 50% for a far-field optical or RF system (near-field can be better), excluding the energy cost of building the transmitter/receiver systems themselves. Losses due to impedance mismatch make this difficult to improve. The most efficient light emitters are ~50% and the best optical photo-voltaics are also ~40%, which would get you to an 80% loss, ignoring atmospheric losses.

> The most efficient light emitters are ~50% and the best optical photo-voltaics are also ~40%, which would get you to an 80% loss, ignoring atmospheric losses.

Unless you have matched emitters and PV cells? For example [1] cites 53.4% conversion efficiency for monochromatic light. (The 40% figure was presumably sunlight conversion efficiency?)

[1] https://ieeexplore.ieee.org/document/4922910

The article makes references to atmospheric conditions:

> "Because of limitations imposed by the atmosphere on the effective transmission of energy within certain sections of the electromagnetic spectrum, researchers have focused on microwave, millimeter-wave, and optical frequencies."

It seems likely that water vapor is the big issue, and that's pretty variable across a lot of regions, and fog and cloud formation would be an additional issue. Hence, this is probably going to be more for relatively short-distance applications (the demo they discuss is a few hundred meters) on Earth. However, in space-to-space applications this doesn't seem to be an issue. Maybe something like solar satellites in orbit around the Moon beaming power to Moonbase would be an option.

For really long-distance energy transport on Earth, the transport of stored chemical or nuclear energy (hydrocarbons or uranium basically, or maybe iron?) are really the only plausible options.

That depends. Is it cloudy? Has a bird flown through the beam path? These are lasers. Everything that can interfere or block a laser will interfere with this tech.

The article mentions a virtual optical fence that cuts the beam if anything impinges, so no roast pigeon with this system. Anything which the beam could couple to, haze, smog, etc, would definitely reduce efficiency for sure.

A cut off beam has an efficiency of zero, so every avoided pigeon is eating into the already meager efficiency of the system. But, good that scientists keep trying, maybe something good will come out of it later.

At these energies they would have to shut down in fog/rain. A single raindrop or snowflake could deflect enough light to potentially blind someone, or at least give them a permanent blind spot on their retina. A trillion raindrops passing through the beam would be statistically very dangerous for anyone, everyone, nearby.

5% per what distance?

> resistance in electrical wires over long distances eat ~5% of the generated power. This makes it infeasible...

Electricity price differences throughout the USA[1] are much more than 5%, so a 5% loss does not (in itself) seem to make long distance transmission uneconomical.

[1] https://www.energybot.com/electricity-rates-by-state.html

Wait until he hears about the efficiency of an ICE. 5% losses for fixed infrastructure feels like not a big deal?

This seems like a very strange comparison to me. Why would you suggest that an ICE is a substitute for moving electricity from solar panels in the southwest to NE?

The percentage losses are irrelevant, except in the context of substitutes. An ICE could be 1% efficient if the next best substitute cost 100x more to operate. Likewise, 5% losses in transmission might add up over very long distances to mean that it makes more sense to build wind in the appalachians or off-shore from NY than it does to power the NE with solar installs in Arizona. OR, beaming energy eating 50% while costing even more due to the amount of exotic materials in construction.

Burning wood might be even less efficient than burning gasoline. But if wood was free, that makes it pretty efficient per dollar.

Yeah, you are not going to fix resistance losses by moving into wireless power transmission.

But keep in mind that the transmission losses are at ~5% because that's a level that people consider economical. They could be higher, or lower. The losses also tend to be around that value for any power grid you look, it doesn't matter if it's a minuscule country or a continental one, people just improve their grid until it gets there.

I'm actually trying to learn more about liquified natural gas right now.

(Eg: if you can gather it sustainably, small batches, can that be used to supplement things other than solar like wood, wind, and hydro? If so shipping that around might make sense.)

I don't know if you'll get many detailed numbers, I got the sense that's the real "business intelligence", not meta level ideas like the one I just expressed, when I was having a series of informal conversations about green energy.

> gather it sustainably

From where?

> small batches

The capex cost of handling equipment ruins this. That's why so much natural gas is already flared.

LNG efficiency in large modern plants appears to be about 80% per https://static.conocophillips.com/files/resources/smid_016_w...

Short answer much much worse than wires. Most of these systems have less than 10% energy transmission efficiency .

Only 5% transmission losses? I don't see how that could possibly prohibit large solar installations in the US southwest. Especially since we already have large solar installations in the southwest.

5% is a fairly trivial loss, especially as PV is usually cheaper than 95% the price of most non-renewable alternatives.

The main limit to long distance, even genuinely global[0], electrical power grid is that it would take several years of current global metal mining to build. But we don’t have to do it all at once, so even this isn’t really a show-stopper.

[0] 5% loss per Mm is 64% loss over 20 Mm with existing HVDC products, and yet because you can place the PV in the globally optimal location where it costs $US13.5/MWh at source and therefore $US37.5/MWh on the antipode, it’s still cheaper. But even for just the electricity worldwide today you need to think in terms of square meters of cross section, and that’s a lot of metal to mine.

On top of that: HVDC scales because resistive loss in power transmission can be reduced quadratically by increasing the voltage (assuming fixed load, so current drops to compensate). (Loss is also increased linearly by transmitting more power through it. Thanks, Joule.)

If your 5% loss per Mm is at 500kV, increasing that to 1MV (entering the territory of UHVDC) means that you're only talking 1.25% loss per Mm which translates to "just" 22% loss over 20Mm, or even just 5% over the ~4Mm that separates Phoenix and NYC.

That tech exists. We just haven't begun to build it in the US. (There are plenty of obstacles beyond just the materials cost.)

I wonder if we might live to see excess electricity being beamed to geosynchronous satellites and then redirected on the fly to locations that need them. That would be pretty awesome even with larger losses and allow better overall land resource usage. Imagine Solar Electricity from the Sahara being beamed to London during a heat wave to decrease the costs of running AC and saving lives.

I’d be very surprised if that works out better than a really big wire, but it’s not inconceivable that a future post-scarcity government may prefer the aesthetic.

I think there's a certain limit where the line losses just from the resistance of the wire would be greater than the dissipation losses from converting the electricity into microwave radiation and reconverting it on the receiving end.

Then again, if we had that technology available then it would make the most sense to have solar panel satellites beaming their acquired energy down to receiver stations rather than generating it on earth and sending it elsewhere.

I wonder how much London, New York, LA, or Dubai would pay for electricity that also cuts 1% of the sunlight overhead while it is being generated, making the weather cooler and decreasing the amount of electricity needed in the process?

It starts interestingly enough talking about optical (which is proven possible though horrendously inefficient), but then goes into snake oil territory talking about charging mobile phones at a distance.

I'm not an expert, so here's one explaining why wireless power is mostly bullshit: https://www.youtube.com/watch?v=MCyLO-1grEk

Note that in all the pictures the receiver seems to be on a tower or at least above the emitter. No doubt this is a safety consideration. Nobody would dare test one of these things horizontally. And good luck with this anywhere near an airport. If they are worried about 1w handheld lasers, a 400w one aimed skyward will require a NOTAM every time you want to turn it on.

The boiling tea thing goes back long before the Japanese team working on this tech. That is a reference to pre-WWII request from the British government for tech that could boil a few liters of water a distance: a death ray. That request eventually resulted in what we now today call radar.

https://www.bbc.com/news/business-41188464

>> "Suppose, just suppose," said Watson Watt to Wilkins, "that you had eight pints of water, 1km [3,000ft] above the ground.

Just put them underground.

At this point you might as well just put cables in the hole that you had to dig up, right?

Wired cables still have electrical resistence.

Better to beam power over buried vacuum tubes.

These systems are less efficient than even mediocre cables.

But if you could affordably construct a continental scale vacuum tube, you could do much better than even a good existing industry standard cable: Vacuum is a good insulator, so you can put any superconducting cable you like in the middle of the vacuum tube and keep it cold with minimal cost.

In principle you could even do away with the cable, by shooting electrons down the tube and collect them at the other end. No resistance so this is a room-temperature superconductor (non-quantum, I believe the term is "ballistic superconductor").

I know, that's why it was a joke.

Buried vacuum tubes would be stupid crazy.

> Note that in all the pictures the receiver seems to be on a tower or at least above the emitter. No doubt this is a safety consideration

My understanding of the article is that the system has a safety mechanism built into it that cuts the beam off when something is detected in the path, so likely it's an issue with providing uninterrupted service rather than safety?

And electrical lines have fuses, insulators and all manner of safety devices. That doesn't mean they don't also regularly start fires and sometimes kill people.

Highly tuned wavelength-specific photovoltaic sounds interesting.

> The receivers for optical power transmission are specialized photovoltaic cells designed to convert a single wavelength of light into electric power with very high efficiency. Indeed, efficiencies can exceed 70 percent, more than double that of a typical solar cell.

I'm still thinking of the last power-beaming example, a Navy microwave-based one, which showed up a month ago (and in this article) & was hyped up but ended up being pretty crazy low end-to-end efficiency if you read the fine fine print. If these folks really can do real 50% end-to-end efficiency- major congrats to them.

At some point, I think solar concentrators probably do make great sense. We already have some pretty big scale solar-thermal plants. These can have nice thermal storage capacity, for off-peak usage. The idea of launching some satellites & trying to point lasers or concentrators down at an on-the-ground collector seems more promising than microwaves, given what lukewarm at best efficiencies we've seen out of attempts to leverage microwaves. That said, the article itself contra-indicates:

> But there have been improvements in efficiency and increased availability of devices that operate at much higher frequencies. Because of limitations imposed by the atmosphere on the effective transmission of energy within certain sections of the electromagnetic spectrum, researchers have focused on microwave, millimeter-wave, and optical frequencies. While microwave frequencies have a slight edge when it comes to efficiency, they require larger antennas. So, for many applications, millimeter-wave or optical links work better.

[1] https://newatlas.com/energy/us-navy-beams-1-6-kw-power-kilom... https://news.ycombinator.com/item?id=31128267 (245 points, 33 days ago, 194 comments)

Cool stuff! I wonder what will be left after the bitter bites of reality for consumer products: consider that we still have actual engineering troubles with things like aluminum in house wiring, which we've been doing for >100 years now.

"battery free cordless" desktop peripherals will be nice, tho.

Wonder what bandwidths the data overlays will be. If you're steering and modulating a beam for power, why not use it for signal at the same time?

Wifi is practical, too. But I prefer wired Ethernet whenever possible, even if it's hard.

Wireless power over radio waves is a dead end. Radio waves go everywhere, great for your cellphone. But completely useless for power beaming. Most of that lost energy doesn't go into heat, but into your body. Lasers are the way to go.

Radio waves, like laser light, can be directed fairly easily.

However, laser light is absorbed by far more things than is radio, which causes them to rapidly heat objects up much more effectively. Some of these things will catch fire, but if the things are water-filled, the effect is much the same for both radio and laser: the water boils until the steam pressure makes them explode.

For example, eyeballs.

No, if you really want safe beamed power, you probably want something that generally doesn’t interact with normal matter at all, but will act with some specific rare thing that isn’t in your body.

Neutrinos would be great from a safety POV, if only we knew of anything that could absorb them usefully so we could turn them into useful energy at the receiving end.

Take a look at the diffraction formula and plug in realistic radio frequencies. Even tens of GHz get all over the place very quickly, not to mention the more common place single digit GHz frequencies.

It’s also reflected off many surfaces, which is how radar works. And IR and visible light bounces off most non-black surfaces (the latter by definition of the word 'black'). UV is also (with strength varying by frequency) diffused by the atmosphere.

But RF needs higher power to become dangerous: Even mere diffuse reflection of 0.5 watt laser light is considered a blinding hazard.

Reflection is not so scary once you have a wave that doesn't go everywhere but really approximates a point source. And if you go high enough with RF frequencies, you'll get significant attenuation when passing through objects just like with light.

> Reflection is not so scary once you have a wave that doesn't go everywhere but really approximates a point source.

Why do you think that helps? Point-like sources make things worse for eyes.

Also, why do you think visible light doesn’t go everywhere? How do you think vision works?

Point like sources can be pointed. That's the whole point. ;) They can be made to not reach eyes, or other places which are not their destination. This is probably very hard to do well, I think there is only one or two companies trying, but at least it is physically possible.

With radio waves, diffraction makes accurate pointing physically impossible. Depending on the frequency, within few inches to few tens of feet, your energy goes all over the place, directly and through reflections.

And the visible light that goes "everywhere" is not coming from a point source, but from a huge omni-directional source in the sky (sun) that passes through a huge diffuser (atmosphere).

I imagine beaming power to space (for powering space devices) and from space (from orbital solar power stations) would be pretty useful.

If you are doing solar anyway and need a decently transparent atmosphere to get power through, why not leave the solar panels on the ground?

Because from a geostationary satellite, you'll collect power all through the night for most of the year. It's like having noon sun for 24 hours a day.

How so? A geostationary satellite sees night time.

Not on most days, because the Earth is a lot smaller in the satellite's sky than it is from LEO, and the equator is tilted with respect to the Earth's orbit. So there's a 44-day period around each equinox, during which there will be shadow for a few minutes per day, up to 72 minutes at the equinox itself. The rest of the time, the satellite gets full sun for 24 hours.

https://www.viasat.com/about/newsroom/blog/how-satellites-ar...

Right-of-way, transmission lines running through expensive real estate, cooling towers heating up rivers, limits to size (you can run out of ground, but you'll never run out of space).

And on the ground you get this annoying eclipse phenomenon called 'night'. Takes away half your potential collecting time.

Also solar flux is like 9X more dense outside the atmosphere. So it takes far smaller surface to collect the same energy.

They had this in SimCity 2000. You'd build these microwave receivers in your city and the orbital solar power stations would beam it down to you. Very clean and convenient. But occasionally the satellite would drift out of alignment and fry part of your city with what you would basically describe as an ultra-high-powered MASER.

Practical designs of this sort of system don't use anything that concentrated - the idea is to build receiver arrays over farmland, since the microwave receivers are really just an array of towers with loops of wire suspended from them. The on-the-ground RF intensity even in the middle of the rectenna would be low - not "advisable to stay in low" but you wouldn't suffer any ill-effects if you strayed in there accidentally for a while while.

The article mentions that in the 1975 they beamed 34kW but it doesn't say why this thread of research was not pursued further.

When is this better than solar panels and batteries?

When your power source is a solar satellite in geosynchronous orbit, for starters :)

Powering/charging aircraft without them having to land.

I could see that working from the top down - ie. something in space beaming power "down" to a plane. Going from the ground up seems like it could be a little risky, nuking other planes, passing birds etc.

Having said that, a Goldeneye-type setup is also a little risky I guess.

I would think it's the other way around: beaming power up, you're less likely to hit people and buildings and terrestrial animals. You'd just hit birds and potentially planes and satellites, but planes and satellites can be tracked so you can shut off power if they're too close.

I imagine that even just powering one receiver from many different transmitters would help you here, since only the receiver would be at the intersection of the numerous beams.

If you want to construct a underground chamber that is electrically isolated from another chamber but still has power. Say for running very sensitive experiments with RF.

How much energy loss has the system? And what complexity curve fit its efficiency over length?

Reply on news.ycombinator.com