Using carbon nanotubes and rectennas to yield a 90% efficient solar cell | Hacker News Reader

Using carbon nanotubes and rectennas to yield a 90% efficient solar cell

pv-magazine-usa.com

268 pointsOsiris30162 comments

This set off one of those "hey, didn't I read about this a long time ago? Like quite a few decades ago?" bells.

A quick trip down memory lane, and here's an archive of one of the stories I was following, back in the day: http://rexresearch.com/marks2/marks.htm. Lumeloid was broadly the same as what NovaSolix proposes -- cheap ultra-high-efficiency PV with optical rectennas -- albeit using polymers rather than carbon nanotubes. I had high hopes that these would be just around the corner in the 1990s, but alas the technology seems to have vanished into the mist. I'm curious to know why: was it a failure of the technology? Financing? Personalities? Alas its demise seems have been undocumented.

Hopefully NovaSolix will fare better; at least today there is complementary infrastructure (far better batteries, electric cars, etc.) which makes investment in any kind of solar tech much more viable than it would have been in the 1990s.

I wonder if the older iteration counts as prior art now?

Outside of the patents for Lumeloid were there any publications?

I skimmed Alvin Marks’ parents. A few things stuck me.

There are a lot of them, like hundreds?

They mostly have one inventor (for something like this, I’m used to seeing teams).

None of the patents I looked at showed any supporting real data. Usually if something is working, early experimental data is shown to support the concept.

They are pretty broad (their are patents for “space train” and “Quantum 3D displays”.

My gut tells me it probably just didn’t work. The basic approach doesn’t look hard to replicate, and I’d expect to see some publications if it worked. So if anyone knows of any publications I’d be curious to see them.

There's a 10min interview from 1991

https://www.youtube.com/watch?v=hMGyob166Cg

crazy to think that so many things were already done and fell below the radar

Plenty of perpetual motion patents, does not mean they work.

The system is designed to protect ideas, not validate them.

Just want to post a reminder that solar cells are already efficient enough and cheap enough to be cost effective for power generation in many regions of the world. I am all for continued research to improve efficiency or decrease cost, but sometimes articles like this one lead people to believe that solar cells are not yet cost effective. They already are!

Do keep in mind that while they are "cost effective" - it takes a few years to recoup the initial investment. A more efficient panel will repay its value sooner, or will require lower initial investment.

Second, the possible applications - Low efficiency cells will support low power appliances, Or you'll need more of them / storage for more power hungry applications. A more efficient cell enables new applications where weight is an issue.

Is there a calculator that can take into consideration the weather over entire year?

Have you examined PVWatts?

I'd also like to know where I can find decent insolation maps

https://www.nrel.gov/gis/solar.html

In terms of power / weight solar can already beat many IC engines.

The only thing more efficient solar brings is less packaging and use in extremely confined spaces like the roof of a car.

PS: The car thing should start taking off in a few years. A Tesla model 3 is 4.6 meters long 1.85 meters wide so 8.7 meters square. At 40% your talking about around 30kwh per day which could really extend the range and for less used cars minimize recharging.

Woah, 30kwh =~ 210 km... that could eliminate charging on even moderately used cars.

Why is weight an issue when we are talking power plants for housing?

Here's a 8.3 kW diesel generator (6.3 kW generator) from Home Depo: https://www.homedepot.com/p/Boss-Precision-Products-Inc-Port...

3900 USD.

367 lbs + fuel = 180 kg + fuel

http://www.nordensolar.se/produktdetaljer/solceller-luxor-so...

One random 265W solar panel, you need 31 off them, one is 18,6 kg gives 582 kg of solar panels to get to 8.3 kW (obviously you wont get 8.3 kW cont. from them). Not counting inverters, mounting rackets etc.

So for equal weight and way more power you have one 8,3 kW diesel generator and 402 kg of diesel.

1 kg diesel is ca 12 kWh energy. Lets guess 30% efficiency gives 1447 kWh electricty for the same weigth. That's about as much as the solar panels will produce in 2,4 months assuming 129 kWh/m2/year.

According to this site it's 3,14 USD per watt solar panel installation for consumers. https://news.energysage.com/10kw-solar-systems-compare-price...

8300 * 3,14 = 26 000 USD.

So, you can buy a generator and 22 000 USD of diesel (about 22 000 litres or 22 000 litres * 10kWh * 0,3 efficiency = 66 000 kWh electricty) for about the same price. Or about what the same dollar in solar panels will produce in 8 years with way lower peak power.

Backpacking is one example where you want lightweight solar power, but aircraft and balloons are a more extreme example. In such cases those stiff panels are a long way from what you want. Also, good luck averaging 30% effecency from a basic diesel generator for home use.

Without going to extreme: https://www.solbian.eu/en/4-flexible-solar-panels

That’s 130 W for 1.7kg. Your 8.3 KW would be 108kg which is 238 lb and thus less than the generator even excluding fuel.

Ye, of course. But the comparison was with ICEs and given that you could bring an ICE weight carrying capacity is usually not that scarce like back packing or balloons.

Depends on if you need power, or, energy over time for the same weight.

Weight is often an issue to some degree. For example the US military has adopted solar for many remote bases because moving fuel is a logistical issue.

Again, I think for the most common case the only real advantage of increased effecency is lower installation costs. I am just saying solar fits in a very wide range of use cases and sometimes say for satilites you want something very specific and paying for it is worth it.

there's an extremely efficient diesel engine called the Lister CS that became popular in the 1920s. It's claim to fame is that it can run 30-45 years without major maintenance and produce 1kwH per .125 gallons of diesel. That's 8kwh per gallon. It was banned in 2009 by the Obama administration because it was considered fuel inefficient by bureaucrats who didn't understand the design. There was a movement to try and de-list it from the ban. My dream has been trying to find one and draft it into freecad so people could 3d print cast iron molds and re-manufacture them. https://youtu.be/NhAD8D3sM2A?t=98

From what I can tell, the Lister was banned because of carbon particulate emissions, not due to efficiency. However, very slow turning engines like the Lister tend to have a more complete combustion and lower emissions.

>>can run 30-45 years without major maintenance

I think it would be better to list the number of miles it can run before maintenance is needed. Just stating the time does not provide a complete picture.

You are right. But let's humor the pitch. What happens if solar gets to 60% efficiency and costs - 20cents per watt?

That solar-powered plane that circled the globe can carry three times as many people ;)

But seriously - The energy density of sunlight is around 1KW/m2 at sea level. So at 60% efficiency you get 600 watt - almost 1 horsepower (~750watt).

That's enough to power a scooter, and with no batteries!

Given that often, cars park outside most of the day, could they charge enough , enabling cars with much smaller batteries(for going to work for ex.) sold meaningfully cheaper - without needing creation of a charger network, which is a huge barrier?

If Sono Motors' Sion lives up to its description, it's approaching usable for commuting using only solar power.

https://sonomotors.com/sion.html/#st-viSono

Edit: Ah, this car is mentioned in the article. Good!

From that site:

> This in turn means that up to 30 kilometers’ additional range can be generated per day in Germany through solar energy

I suppose that 30km does cover a lot of commutes (15km each way), although "up to" is a notorious weasel word.

Based on the numbers quoted on the site, 30 km of range is equivalent to what you'd get if all of the solar cells got 4 hours of direct illumination.

On the one hand, they can't all be in direct sunlight at the same time, given the car's geometry. On the other hand, many locations get more than 4 sun-hours per day on a horizontal surface[1], especially in summer. So it doesn't seem too implausible.

And if your commute is 30km per day, then even if you don't get enough sunlight to reliably charge the battery every day, you still have more than a week of buffer in your battery.

[1]: https://www.nrel.gov/gis/images/solar/solar_ghi_2018_usa_sca...

I think you can buy panels as extras for existing cars, they provide a couple of miles a day under the best circumstances. If it was 10 miles under best circumstances and a couple if left anywhere outside, I could see them becoming mainstream.

It could work, depending on your use case. For someone who only drives a couple miles a day and parks in the sun, they might need to charge less often or not at all.

If you're going on long road trips, it probably wouldn't help much. In that sense, solar isn't particularly useful for extending range (which is why you'd want a large battery in the first place), but it might make a big difference when it comes to how often you have to plug in to a charger.

Still could be good if the surface area available is limited.

Just to footstomp this, there are other costs that are already beginning to dominate the analysis for solar, like customer acquisition costs. Convincing someone to attach something to their house has significant costs. Like, "over $2000 per install" significant.[0]

Other soft costs, like land acquisition and taxes, still play dominant roles in many projects.[1]

The naive assumption is that solar succeeds when there's lots of sun and it's all just physics.

But these other factors, like, say opportunity costs (the cost of available substitutes), are why you're often better off installing solar in Alaska than Florida.[2]

Theoretical physics are great, but at this point I'd be far more excited about big ideas to deal with the messy economics side of solar generation. An announcement that some sort of "solar council" was going to pursue a national marketing campaign to drive solar adoption would be huge news. Or Elon deciding that strip malls were his new target, to drive installation of solar on every commercially leased roof, which would reap huge efficiencies of scale in local installation costs over residential.

Or, the cheesy TED talk clincher would be: the next big solution isn't always about electrons, sometimes it's about people.

[0] https://www.solarreviews.com/solar-leads/customer-acquisitio...

[1] https://www.nrel.gov/docs/fy17osti/68925.pdf

[2] https://solarpowerrocks.com/solar-basics/how-much-electricit...

OK, can you just buy a solar heating panel? You and your best friend can install it...depending on location you can easily get hot water for 6-10 months a year. That's quite a savings. You don't have to solve everything today ;)

Solar heating panel? A bunch of wooden/plywood boxes, some glazing for them (can actually reuse thrown-away window panels), a bit of plumbing with PP pipe (ridiculously lego-like easy), some kind of top tank (a common steel drum?).

And some making sure the roof/attic floor doesn't collapse, which usually is not a problem.

This can be _build_from_scratch_ in a couple of weekends with a best friend.

But for the construction codes. And unions.

It's so low-tech, so I wonder why I haven't heard of 19th century installations. Me, I actuially built two in places where construction codes are not an issue. And have seen functioning installs dating from 1950s.

I don't think it even makes any sense to mass-produce heat boxes the way the solar-electric panels have to be. It all can be either made on site or included into the house design from start.

That's maybe why you can't buy solar heating off the shelf.

I don't agree simply because it is very common for diy solar water panel systems to leak all over the place. There at least needs to be prescriptive code to help reduce instances of crap work. But since it is a lot of pipe in a small footprint, it just makes sense to have it made in a factory with better quality control.

You and your best friend most likely can’t install it. You not only need building permits, you also typically need a licensed electrician to do the work. Doing it without a permit or properly credentialed installers risks sanction from the municipality as well as potential insurance consequences.

A 'solar heating panel' in most cases is a water tank exposed to the sun as much as possible painted in black. Or maybe put inside a black box. Not a lot of electricity work there.

It's not quite that simple. Flat plate collectors are roof components, and they have all the problems of roofing. Making them leakproof despite big thermal stresses takes some work. Brazed copper plumbing works. Plastics, not as well.[1]

You usually want a water tank to buffer the thing, so you can get hot water at night. Temperature will be all over the place without that. Giant hot water tanks in attics need extra structural support, which was a problem in Florida at one time.

There are some fancier setups that will work in cold weather. Some systems use antifreeze in the panels, and a heat exchanger.

[1] http://www.alternative-energy-tutorials.com/solar-hot-water/...

Sort of true. It is indeed the case that, at least in the US, the cost of the cells is a minor part of the cost of a rooftop solar array. You’re missing one detail, though: a lot of the costs you’re describing scale with the area of the array — the costs of mounting, wiring, sealing roof penetrations, shipping, physically hauling the panels to the roof, etc. are all related to the number and size of the panels. If you can shrink the array by a factor of 3, it makes a big difference.

I’ve read that the economics may already be close to favoring multi-junction cells for this reason, and they’re nowhere near 90% efficient, nor are they cheap.

At some point, there has to be a standard to which roofs are built (including wiring and power ports) such that you can just buy shingles and put them in or replace from the inside. Just cut the middle men and installation costs.

I hate to be a bummer but based on the way the construction industry works, that's almost certainly not going to happen. For example, in the U.S. most plumbing jobs fall under a bevy of regulations and yet tasks as simple as changing a faucet can frequently lead to issues like unmatched pipe diameters or thead types. We've had 100 years to standardize that and it's still a mess. You literally can't even just go buy a garden hose and know for sure it's going to fit your exterior faucet.

Electrical is arguably the most regulated, standards based and straight forward aspect of construction and even that varies a great deal based on the electrician and equipment.

A rigorous standards based construction industry would be amazing. We could design electrical systems that could be installed by a novice with almost no risk of danger or failure. Homeowners could change luminares just plugging them in and attaching them to an electric box toolesly. We could have plug and play water appliances.

Damn it would be magical but again I'm fairly certain that's just not going to happen.

Edit: Above I indicated that the construction industry doesn't operate on "rigorous standards". That's not really true. There are actually very good science based standards for most construction tasks. It's more of a problem that the industry doesn't really focus on making future work on a structure low cost. And it makes very little effort to increase accessibility by creating more consumer facing equipment and tools.

This is, I assume, because standards and regulation are aimed more at safety than at ease of maintenance/modification?

The majority of construction regulations I've run into are relatively flexible. It's expensive but if you want to try out a new material or technique, the path to validating it and getting it approved is well defined. The problem is that few construction companies operate at that kind of scale and outside of churches, flagship venues (i.e. Disney hall), and palaces there just isn't enough money to develop new technologies.

That says a lot about the US. In Europe, you can buy only one type of garden hose because there is only one kind of port. If the EU would introduce a new standard to which roofs have to be built, then people will start using it within 3 years. And they will benefit a lot from that standard, so you will think twice before buying a new roof for 20k+ and not buy from the companies that keep to this standard. Also don’t underestimate the power of adapters.

I think you also conflate the way which something is built with the specification it is built against - like interfaces vs implementation details. If you tell companies to build a certain kind of roof, with ports and wiring, then they will not receive any money if they don’t meet your spec. They are free to use the kind of wood, nails, wires (as long as thick enough) that they like, and this is where they can compete against each other.

Perhaps the parent comment has had some bad luck, but I've lived in a half-dozen US states, and have never seen any garden hose tap that was different than the standard set in the 60s. FWIW. https://en.m.wikipedia.org/wiki/Garden_hose

Garden faucets in the U.S. can be 3/8, 1/2, 5/8 or 3/4 inches. They can have a course thread or if a backflow preventer is reuired in your muni, they have a fine thread.

I've seen hoses with all of those diameters, but they've all had the same standard fitting size at the ends. Same with the backflow preventers- all the normal course threads. Now I'm genuinely curious, what state are you in? (not being obstinate, I really am just curious!)

No offense taken. I'm in Western Washington state. The home I previously lived in had fine machine threads on the exterior faucet. I assume this was to prevent attachment of a hose without a vacuum breaker. Here's the item I had to purchase to attach a standard garden hose.

https://www.amazon.com/dp/B002OU6ZSA/ref=cm_sw_r_cp_apa_tbA-...

Now that I think about it though I believe you are correct about the diameter of the port being consistent. It is indoor plumbing valves that I've run into a variety of sizes of. 1/2" vs 3/8 compression.

Interesting. Thanks for answering!

This is not true. Although Gardena has a huge market penetration in the EU, it is not a EU standard.

There are other types, no EU involvement here ( yet ).

Did I say that garden faucets are standardized? Yeah you can get other sizes than the Gardena ones I guess, but those are more obscure. I think you will agree though that EU standards are usually followed.

Simply not true. In Poland we not only have two standards, they are also expressed in inches: 1/2 and 3/4.

Yes, but only two and they are a no-brainer because of adapters.

If you can design a roof from which a shingle can be replaced from the inside, the roof doesn't leak when exposed to wind-driven rain, the roof can hold your weight, and the cost isn't hilariously high, I'll be impressed.

Also, if you stick a power port on your roof and you live in an area with lots of lightning, I'll stay away from your house.

Basically every window is opened and can be replaced from the inside. How is that different?

Regarding lightning, IDK how the Tesla roof solved this, but they obviously did.

Have you ever seen a roof built? There’s a roof deck, and shingles of some sort are nailed in from above in a specific order. In many installations, there is a weatherproof barrier between the roof deck and the shingles. If you want to replace a single shingle in the middle, you do it carefully, and you may need a special fastener.

That’s how they are now built, yes. Cars looked very much different 100 years ago, too.

Another issue is theft. I own a remote desert property in many ways ideal for solar, but the utility electric bill averages < $20/mo.

Nobody is going to try steal the hot copper wires connecting my house to the grid.

But where there's a substantial PV array on display, there's batteries and an expensive inverter nearby, and not much to prevent taking them.

Yours is a very unusual case. Most domestic solar installations are on roofs, so not likely to be stolen, and most homes' electricity bill is much higher than $20/mo.

I don't expect the panels to be the first things taken.

The battery bank and inverter however, which are easily put invisibly to use by anyone even your neighbor, are quite the prize.

I'm not sure it's all that unusual a case. Solar is especially relevant to remote off-grid homes. Unfortunately, these are also prime targets for theft as they're often left unattended for long periods with distant neighbors.

Denser populations have a grid, and can just have a large solar installation powering that grid on a large scale.

It's a problem. Solar puts expensive infrastructure on-site, it's a mini power station.

> I'm not sure it's all that unusual a case.

> Solar is especially relevant to remote off-grid homes. Unfortunately, these are also prime targets for theft as they're often left unattended for long periods with distant neighbors.

Remote off grid homes are a very unusual case. Not saying there is no demand from that segment, but even there is a higher rate of demand from it, it's a small segment of the market vs urban/suburban rooftop solar, and I wouldn't expect major manufacturers and installers to optimize for it.

> Denser populations have a grid, and can just have a large solar installation powering that grid on a large scale.

That's already happening today. However people choose to have rooftop domestic solar for many reasons ... to fix the long term cost of their electricity consumption, to offset peak power rates, etc.

> It's a problem. Solar puts expensive infrastructure on-site, it's a mini power station.

In situations where site security is a problem, presumably custom solutions can be put in place by installers to ensure security of the inverters and batteries. It may cost more, but that's sort of the price paid for owning property in a sparse, remote location.

Also in some parts of the world they will definitely steal your copper wires (out of the walls of your house if there isn't sufficient security!) But in those places your solar gear wouldn't last a week.

In the UK an acquaintance of mine has regular phone outages because the copper comms lines are pulled. They're shallow ditch laid; the thieves hook one end to a tow-hitch and drive, I gather.

I've heard about this happening, fortunately it's not that bad where my property is. Not yet anyways...

Are you seriously arguing deployment should cease because of a theft risk?

I don't doubt the theft risk exists. I'm asking if you are catastrophising to "..so we shouldn't do it" or just observing a downside risk component.

Nope, not at all, I wasn't aware I was engaged in an argument.

It just makes more sense in a lot of cases to let the utility provider convert to solar than turn your house into a power station you then have to secure accordingly and maintain.

I'm not interested in having to surround my property with barbed wire topped chain link fences, which is what you'll see around power stations for good reasons.

Mostly risk of death reasons. Solar and battery are not high on my list of lifestyle and risk components any more than a car parked outside or plants are.

Maybe you live in a much higher crime risk area. I know (for instance) honey hives in NZ are being predated by gangs so I can believe there are portable high theft risk things in rural settings. But I'm really unsure a Tesla battery rig is one.

I've had plants stolen. I still maintain my garden..

> I've had plants stolen. I still maintain my garden..

What a world it would be if independent solar power were the cost of a garden!

How much do you value your time? Gardens tend to be pretty expensive when you add up the hours for a decade or two!

Ah yes.. economic efficacy meets relaxation and hobby. You can't buy pre-read books quite the way flann O'Brien wrote about but it feels like we're close.

Having said which, close friends ran bespoke gardening for richer garden aesthetes.

Back to the lead topic. PV is worth doing but assess your theft and vandalism exposure in risk/cost benefit.

Data, please: how many documented cases of this type of theft actually exist? No offense, but it sounds like you live in the middle of nowhere and are slightly paranoid.

Rural theft is a big problem around here. Just a few months ago someone stole the entire generator at a tower site that is high up on top of Grey Butte, near Smith Rock in Redmond, Oregon. I picture a similar risk for solar panels.

Yes. Rural theft is a problem. Now, explain how society advances if we give in to crime and we stop investing in rural communities

>Nobody is going to try steal the hot copper wires connecting my house to the grid.

Especially considering the wires connecting your house to the grid are actually aluminum.

Is PV theft really a problem though? I mean they could sell them on ebay but AFAIK most thieves aren't just fencing everything on ebay. I can't imagine a pawn shop would give them much for it considering how hard it would be for the pawn shop to resell them. With copper and aluminum theft the thieves can just go down to the nearest scrapyard and get paid by the pound.

> Or, the cheesy TED talk clincher would be: the next big solution isn't always about electrons, sometimes it's about people.

An even cheesier TED talk clincher would be:

> the next big solution isn't always about electrons, sometimes it's about elections.

Plug and Play Solar is one way to dramatically lower soft costs of residential pv installs. Fraunhofer is working on this [1] and there are already other 3rd party products on the market.

[1] https://www.cse.fraunhofer.org/pnp

"Convincing someone to attach something to their house has significant costs. Like, "over $2000 per install" significant."

Not any more. We've got 50-state legal solar stuff where you just lay the panels out in the yard, literally plug them into your wall outlet, and every outlet on that circuit is energized. You can even leave it hooked to your power at night time, as the junction boxes have barrier rectifier diodes that prevent the panels from consuming power at night.

Install costs are becoming almost negligible due to advancing technology.

Cell efficiency is up there. Typical monocrystalline panels run 20+% now days (and look utterly gorgeous - https://imgur.com/a/sZGtYmB - picture from my job.)

Opportunity costs? Each cell runs on average $1.00 for most manufacturers, so you just need to shop around. A typical 60-cell panel at 300W runs about $150-200.

Most of your concerns have not been concerns for about 5 years.

Plus the environmental costs of creating and, at end of life, disposing of the arrays.

You’re a good thinker and writer. I’m definitely ready for your TED talk!

Honest question: do you think residential / small scale commercial solar is the future? As an outsider it seems like utility companies are best placed to deploy solar at large scale in the right locations, bypassing customer acquisition costs.

Here in California, PG&E offers the "solar choice" program where I pay extra to have 100% of my electricity generated from solar. This cost $6.50 of my $82 electric bill last month. (Yes I know electricity is fungible.)

That's a tough question.

Distributed production offers some huge advantages, but in the nrel study above you'll find that utility scale is strictly more efficient.

I think Hawaii is a good case study. Customers are generally frustrated with the power company's poor reliability and high rates. Rooftop solar offers a substitute for what has traditionally been a protected monopoly, and there's abundant sun, so solar adoption there is crazy strong. HECO has generally fought to slow it down.

If utilities realize this threat, they'll become competitive on price and customer service. Reliability in some places.

If they do not understand this threat, then they will just try to lobby for more and more extensive penalties for those with rooftop solar who are still connecting to the grid. Like cable companies raising rates in response to the rise of Netflix.

The most extreme version of that would be the emergence of microgrids - cord cutting from utilities by entire subdivisions.

Microgrids are potentially really cool technology, but would require some pretty intense dysfunction to take off.

So while I hope utilities figure this out, regions will vary, and rooftops put a new sort of check on how bad utilities can get, even if they don't become ubiquitous. Rooftops will inevitably be the answer in some places, not all.

Oh and if you look country by country, to places where it has been harder to get a well financed responsible utility... Parts of Latin America or Africa could really be jump-started economically by rooftop solar.

I did an analysis of coats for residential roof top installations in Europe for my Master Thesis. Overe here around 30 to 35% of costs are the panels. Still the most expensive part but nowhere near the 60 to 70% it used to be. Panels really turned into a commodity.

What struck me is that nobody, not even Solar City and others, looked at the logistics side of things. It seemed to me that eveeybody came up with elaborate and capoztal intensive ways to sell to the people with the smallest customer Life-Time value in the whole value chain. Putting sophisticated logistics in place to cut costs (my assumption was that by doing that the solar panel cost could be reduced by around 15-20%) and selling to installers would have been the better way to go.

So I second you and the OP, it is no ponger about physics (with the exception of large power plants) and more about people and how you get solar on said people's roofs.

Is your thesis in English and avalaible online?

German, unfortunately I couldn't convince my Prof to accept it in English. In case you're interested and not under time pressure I can translate it or part of it. Always happy to share stuff like that. Just shoot me a mail. hef19898 (at) gmail.com

Sorry for the long delay, but would encourage you to at least translate the key parts into English.

Started already and managed to get to page 12 of 78 when the people realized at work that Christmas is, surprise, again on Dec. 24th. I totally forget that I wrote that many words... ;-)

>other costs that are already beginning to dominate

Which is an argument for trying to make cells more efficient like the ones in the article apparently looking at 43-90% as against something like 18% for typical cell. If most of your costs are installing etc then those can remain the same while you get twice the output.

Yes, and also making that house, or land, or taxes produce 4x as much energy is categorically different form reducing the cost of existing 22%-efficiency panels by 75%.

That's because higher-energy-per-unit-area panels increases the efficiency of land use, which drives all the other costs you list: land acquisition, taxes, customer acquisition, etc.

The real breaking point will be when solar is so cheap that it costs less to replace a coal or natural gas plant than it would be to continue running it.

Well you also need to solve the storage problem. Your solar plant will not replace the coal or natural gas plant at night or on cloudy days.

Just convert electricity to hydrogen.

Yes, actually it already is cheaper to build new solar or wind than run existing plants, at least in some situations.

As you say intermittency and time of day is the issue. Although I think we should just reflect lower costs in time of day tariffs. I suspect there are a lot of use-cases where people would happily shift demand to save money. For instance if you could just put your clothes in a washer dryer, and leave it to handle the cheapest time. Or air conditioning which creates ice earlier in the day then uses the cache later. Provide the tariffs, open the market opportunity, and see what technologies emerge.

So you agree with GP that the storage problem is currently a problem and needs to be solved.

Controversial statement alert:

The storage 'problem' doesn't really exist. Do we 'store' electricity currently? No, we just generate it all the time and it's 'used' straight away. Well, what we do technically now 'store' will continue to be stored if using wind/solar (e.g. pumped hydro 'Electric Mountain' in the UK).

Shall we also pedantically say we current 'store' electricity in the form or chemical energy in gas/fossil fuels or whatever? We can use that definition I suppose.

Even with that definition no leap to any new 'storage' method actually needed if we look at the functional output required to keep the system operating the same way it does now. You just need need to have enough renewable over-capacity to generate enough electricity to meet the 24h cycle. Which is functionally what we do now.

The "it's always windy/sunny/tidal somewhere" solution is the same as what we do currently, we just put our hands over our eyes and ignore the 'energy security' implications of not producing enough oil/gas to meet our own needs so it feels like somehow what we have now is more certain.

That's not to say it wouldn't be more efficient to store electricity and release it as needed. But just like now what would solve the problem is over capacity with enough resilience to feed the 24 hour cycle.

Can that be done with 'just' wind and solar? I don't know,I guess not. Nuclear would seem a good addition to the mix, and tidal or wave power.

But it seems like we should at least approach the problem openly without hamstringing ourselves worrying about 'storing' electricity being something we absolutely must 'solve' before renewables can be relied upon.

> No, we just generate it all the time and it's 'used' straight away.

Right, and we vary how much we produce to match demand. We can't make it sunnier just because everyone has gone to put the kettle on at the same time.

I think the point is that the existing grid has the ability to do that on-demand scaling, so adding variable solar or wind capacity to the grid means the problem is already dealt with. Of course as we transition to higher percentages of solar and wind capacity the demand scaling will become a bigger issue and storing the solar and wind output will be needed. I'm not sure how far away that is now - perhaps the existing hydro and nuclear capacity is enough to deal with all of that variability, but I expect at some point we'll need to figure out the storage issue.

We vary how much we produce by having over-capacity, and using a portion of that to 'turn up the generators' as needed.

If there wasn't enough generating capacity in the fossil fuel system it would be equivalent to 'can't make it sunnier'. The flip side to 'can't make it sunnier' is you can add more solar generating capacity (over-capacity) and add mixed generation through wind/solar/tidal. Both options end up at the same place - over-capacity in order to meet peak/daily/yearly loads.

Classic fossil fuel generating stations quite literally sit idle waiting to go, that's the over-capacity. The same thing can be done for renewables (the whole mix, not just solar).

And to answer the inevitable 'but we've always got access to fossil fuels, like at night when the sun doesn't shine'...

Right now we 'store' the electrical energy in fossil fuels (really we release it) and say this system if inherently more stable because we assume access to those fuels is immutable. And we say that the energy 'stored' in the wind or tide or sun is not.

In reality, fossil fuel energy security is a delicate balancing act of just in time logistics and political power plays, for most nations without vast resources or nuclear generating capacity for instance.

Functionally, perceived limitless access to fossil fuels gives the false impression of superiority and conversely give the the perceived absolute need for renewable storage because of the very obvious daily cycle of light/wind/tide.

I simply say that adopting a system of over-capacity like we currently have for fossil fuels and recognising that neither renewable/fossil fuel resource access is certain is a sensible middle ground upon which to build a sensible generating program. Both systems can rely on having more capacity to handle peak loads, both systems have the ability to plan for their faults and both can work without massive grid scale storage.

Adding storage to renewable generation would actually be a huge step up over the current system and would be fantastic. I would love to see such a system work, and I believe it will do anyway through distributed smart grid/EV usage and large grid scale battery storage.

It's just the uncertainties of the current system are hidden from the average user and I'm just saying we're setting a higher bar for renewable use and storage because the access limitation are that bit more obvious on the daily cycle.

Sure, but the reverse side of that is the proportional cost increase of solar that has excess capacity. Solar has negligible running costs, it's irrelevant that you use the energy or not, the capital costs will have to be recovered from the price of the fraction of energy that you do use.

So you are effectively asking for a 3-5x lower solar cost before it could replace existing sources. Never mind that there are seasons and periods of the day when no amount of excess capacity will cover the demand, there is simply no significant solar energy on continent-wide scales.

Everyone keeps talking about solar. Probably for the classic 'sun doesn't shine at night' thing.

I've repeatedly said to use a mix of renewables, just like a mix of fossil fuels is used, so deficiencies in one is balanced by another.

The exact same thing is done in traditional generation. You don't use a nuclear plant t for a TV pickup for instance, you wouldn't use solar for the whole grid.

Balancing the way you generate reduces the amount of over-capacity required (and subsequently the cost), exactly as is done now in traditional generation.

We definitely need more in the way of cost-effective energy storage. There are only limited places for pumped hydro, unfortunately, and battery tech isn’t yet cheap enough over its lifetime.

I’m optimistic, but we’re not there yet.

This is a solved problem: https://qz.com/1355672/stacking-concrete-blocks-is-a-surpris...

Also hydrogen.

We use the word "solved" differently.

“A 120-meter (nearly 400-foot) tall, six-armed crane stands in the middle. In the discharged state, concrete cylinders weighing 35 metric tons each…”

120 m * 35,000 kg * 9.8 m/s/s = 41.2 MJ = 11.4 kWh.

LiIon: 250–693 W·h/L, 100–265 W·h/kg, 250~300 USD/kWh in 2018

-> 11.4 kWh = 45.7-16.5 litres, 114-43 kg, $2850-$3420.

Concrete is much harder for me to find number for, seems to be $92/cubic yard, 2.4 tons/cubic meter.

35 tons -> 14.5 cubic meters -> ~19 cubic yards -> $1748 plus whatever the cost of a crane that can lift 35 tons is.

My knowledge of cranes is whatever came up first on Google:

“””A Comadil CTT 361-20 is a good size crane able to do concrete panels, max jib length is 75 meters and they have a max lift capacity of 20 tonnes. These cranes can go up pretty high without needing to be tied into the building. This crane will cost around $1 million”””

Article later talks about renting smaller 4-5 ton capacity cranes for $1500-$2000 per week.

The problem is neither LiIon nor concrete with cranes nor hydrogen are cheap enough. Not yet.

It is only a problem with current solar prices. With much cheaper solar one can produce, say, methanol and store enough of it to generate electricity for months. And that will be cheaper than burning natural gas.

Yes, I agree it’s a problem, or rather a series of problems. Each one you solve allows the percentage of renewables to increase. What I’m saying is you need to set up the market opportunity as early as possible so that people can start working on commercial-scale solutions.

From last week, albeit for wind:

https://www.cbsnews.com/news/its-now-cheaper-to-build-a-new-...

Existing coals pretty expensive these days, and that's not even counting its high costs for health and carbon. For example, Indiana is retiring tons of coal and replacing it with renewables to save money:

https://www.utilitydive.com/news/even-in-indiana-new-renewab...

I only wish we'd have less ~toxic PV cells production. So far silicon requires CO2 and chlor- compounds.

Anyone with 20%~ efficiency and a clean manufacturing process is rich for the next millenium.

You do realize that table salt is a "chlor- compound", right?

Yes, refining silicon requires the use of dangerous chemicals. No, that does not make it "dirty" in the way that burning coal is dirty, i.e. expelling carcinogens into the air.

I know but still.

No, not “but still”. Water is one oxygen atom from being bleach. Chlorine is metabolically essential.

Chemistry doesn’t care about buzzwords. If you don’t specify what exactly is bad about the industrial handling of whichever chemicals actually are toxic in this case, you’re just flag-waving, not helping.

what buzzword, the cleaner the process the easier it is to produce anywhere, the easier it is to use solar instead of fossile fuel. I just forgot the actual name: tetrachlorosilane when I typed my first comment

To avoid this http://www.washingtonpost.com/wp-dyn/content/article/2008/03...

Um, you do realize that silicon fab lines generally recycle practically everything in terms of waste materials?

It used to be that the waste material from processing was more pure than the actual industrial infeed material. So, the first thing that semiconductor fab lines would do would be to take their "new" industrial material and feed it into their "used" line for reprocessing. I think this has changed as the industrial sources have gotten better, but I suspect that recapture is still almost perfect.

No I didn't know that, I read a few pages saying waste materials weren't recycled hence my comment.

This misses the investment context which leads to companies choosing coal oil and gas over solar.

The limiting factor on deployments continues to be "what to do when the sun doesn't shine bright". Night for starters and in a lot of the world, you have whole seasons where the skies are overcast nearly all day (monsoons).

If you anyway have to couple solar with something else to make up for the variability in power generation (largely fuel based because Hydro and nuclear are massive government led projects) and you have to invest in expensive and bulky power storage to handle even the nights, why not pay a little more incremental money and get a slightly bigger fuel based power plant.

It's also important to realise that, in any given situation 'cost effective' doesn't necessarily mean 'affordable' (or affordable-enough to lead to the people installing them).

The people need to have the available funds, etc. It may save momey but that's not an immediate payoff.

Any increase in efficiency vs costs will help with that equation.

Improving efficiency means lower footprint per kW of installed capacity. People are now viewing solar in terms of area of installation, contract with power company, etc.

Even in rural India, the idea that solar is cheap has crept in. Farmers are buying solar power pumps across many states.

It's not just about costs. It's about size!

The higher the efficiency, the less solar cells you need.

This has impacts on environment (solar cell farms), adds possibilities for miniaturization (what, if a vehicles rooftop could feed all the energy needs of said vehicle or, at least, constantly refuel it?), mobile device chargers (could be even become the outer hull of a battery, so you just need to place the battery in light to charge it), small cells, that produce high energy, could be made part of clothing (must not be this tech, I am thinking about miniaturization in general). Possibilities seem endless...

Here's the money shot:

https://l0dl1j3lc42iebd82042pgl2-wpengine.netdna-ssl.com/wp-...

If they can really make that happen cheaply (green shaded areas, meaning efficient collection of power from near IR photons) this is revolutionary.

Current silicon based tech (blue shaded) is useless for light energies below red.

I am always fascinated by how people that don't waste their time with the latest data visualization trends can come up with slides so ugly according to latest trends yet so effective at exposing information.

Apart from the (ugly) colors, the plot looks like standard R/gnuplot/whatever tool existing since (I grow old, I grow old) the beginning of time...

My fascination is more at how dysfunctional "prettifiers" of graph manage to turn useful data into.

That graph is perfect. Axis labeled, legend present, shit is legible and has pixel resolution. Way to go!

I'm surprised that the nanotube antennas seem to capture all frequencies equally. I would have expected resonances and nulls.

I would imagine they rely on the nanotube lengths being non-uniform. Manufacturing variability as a feature!

Folks here are missing the fact that the rectifying diodes must operate in the THz region. The fact that NovaSolix has demonstrated Proof-of-Concept chips says they are able to build diodes that work at 100s of THz. That, alone, seems impressive to me. (example: green is 630 nm; f = c / lambda, that is, 476 THz)

Just to check my understanding:

A green light would create a AC current at 476 THz. We put this through a diode to capture the pulses of current that are moving in the right direction?

Yes.

Also, the voltage won't be very high. The diodes have to be efficient in the millivolt regime.

yours is a really good question, one of the links answers it:

https://today.uconn.edu/2013/02/uconn-professors-patented-te...

Look at the picture of the electrode with triangles and the other rectangular. The distance between the tip of the triangle and the other electrode is required to be on the order of a nm (in this reference here 1.5nm). This allows electrons to tunnel from one electrode to the other. A charge on an electrode tries to spread out, and the density is higher near pointed protrusions. So in the cycle the triangle-electrode is negatively charged, the electrons can tunnel to the other electrode, but in the other half of the cycle the electrodes prefer to spread out differently on the rectangular electrode. That's how the diode is implemented.

Edit: just adding, so the asymmetry of the diode is due to the asymmetry between electrodes and the asymmetry between electron and hole mobilities

A carbon nano tube antenna that is capable of working at infra-red frequencies have been around a few years, thats fairly simple[1]

The hard part was finding a diode/rectifier that was capable of running at terahertz frequency. I cant see how this rectifies. I assume its to do with the layout: https://l0dl1j3lc42iebd82042pgl2-wpengine.netdna-ssl.com/wp-...

If they are not telling porkies, then this is a wondrous breakthrough. (assuming the nanotubes are UV stable.)

[1] 1/2 wavelength conductor, which is possible with carbon nanotubes

90%? I thought the theoretical maximum was 86%: https://en.m.wikipedia.org/wiki/Thermodynamic_efficiency_lim...

https://l0dl1j3lc42iebd82042pgl2-wpengine.netdna-ssl.com/wp-...

This one shows 85% efficiency? :)

I think you are on to something. For non-concentrated solar, what a low-cost application would use, the limit is given as 43%. The numbers given trigger my bull-detector hard and I've had some insight into the difficulties that you face if you actually want to build a rectifying diode. Those things are hard. Really hard. You're getting nowhere near where you'd have to be to achieve the limit. To satisfy the claims given, these guys have to be geniuses beyond imagination.

If something sounds too good to be true, it probably is.

> And like so many other ideas, if the sales pitch is real, if the scientists can deliver, if the hypothesis can move to theory – and the theory become applicable in a scalable, manufacturable good – then all the rules change.

Wow - this is a totally different approach than crystalline silicon, and could be much cheaper to produce as well as more efficient. Something new is going to be required to do much better than 35% efficiency, as you can only do so much lining up the band-gap of semiconductors.

On another note, it's nice to see that crystalline silicon PV is perceived to be so successful that people are lining up to comment on how a dramatic improvement in power density and cost wouldn't really matter.

Carbon nanotubes, the wonder material that's only as toxic as asbestos [1].

[1] https://www.scientificamerican.com/article/carbon-nanotube-d...

I assume this won't ever hit the market at the promised performance and cost. In the past decade or so we've been inundated with hundreds of stories, maybe thousands, about new solar breakthroughs and they never come to pass. The same is true of all the battery "breakthroughs". For some reason this stuff is never commercially viable. I wish I understood why.

Just a random thought on the idea that there might be a limit function as it pertains to the desirablility of increased efficiency.

We have a 13 kW array. I make it a point to keep it cleen. However, this isn’t something one can expect to accomplish daily. At the very least it would waste a lot of water. As a result of this, my panels, on average, enjoy a combination of dust, leaves and, yes, bird poo, sometimes lots of it.

If we had 90% efficient panels the combination of dirt, dust and a determined flock of birds could take out 50% of my energy generation capacity. Having a less efficient system distributed across a larger area might actually be a better idea.

90% efficient panels would have ~4x smaller area. Wouldn't be that easier to clean and secure from bird poo etc?

I don’t feel like getting too scientific today. My gut feeling is that the accelerated loss rate due to greater Watts per square meter will at some point intersect with the curve denoting cleaning needs. At one point water waste and man-hours might become a problem.

Their argument that it will be cheaper because of raw materials costs is pure rubbish. That was supposed to be the point of silicon - it’s just sand! Then people realized that the structure (crystalline) and energy intensive processing required to obtain that structure were the real cost drivers, not to mention the housing, transparent electrode arrays, and installation costs. Nanotechnology is difficult to manufacture because of the energetic factors working against the formation of long-range nanostructure.

Energetic? It sounds like it would be entropic. Every crystal or polymer is a long-range nanostructure.

Sorry, you’re right- but entropy is a part of the thermodynamic free energy of a structure, I was glibly rolling all that into “energy”..

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

We can use simple processes to create crystals and polymers for only slightly more energy than the Gibbs energy of the resulting structure. Currently, fabricating custom nanostructures requires vastly more energy than that. A resist + lithography + etching process has to manipulate hundreds of atoms to get each bit of entropy.

Biological systems are efficient at creating nanostructures. That may be the future of fabrication. It'll never be repeatable enough for silicon structures like DRAM, but it might someday work for bulk surfaces like solar cells.

I think scientists and inventors often do themselves no favors by making claims like "90% efficient". That's the theoretical maximum. Now if they hit 80%, we'll all say they failed despite it being an incredible achievement.

They should claim "75% or more", and then hit 80% and blow our minds. The Scotty Factor[0] should be liberally applied when possible[0].

[0]http://wiki.c2.com/?ScottyFactor

The more important number is .3cents/KWh a 1/10th the cost of current prices. Also, they don't use the exotic materials.

I don't think that's an issue at all. If the panel actually comes out and it's 60% at comparable prices to existing ones I'll still be overjoyed.

The highest cost of Solar installations these days is the installation cost itself and the batteries -- if you're willing to spend the cost for the batteries. 90% efficient is great but the real value for most would be in reducing the cost in the additionals to the solar cells.

Less than 40% of the installation costs are the actual solar panels.

If someone came out with a DIY solar installation, that would be a major advance.

I've done numerous DIY solar installs, to power outbuildings on my property, as off-grid battery-powered systems. I hired professional help for our system that powers the main house, as the scale was larger and the grid-tie was beyond my expertise. But an experienced electrician could surely do that part as well. So the lack of DIY installs is less about the solar tech, and more about where any given individual's DIY skills fall.

There's a speculative solar announcement like this every week. Until they actually produce results don't get your hopes up.

Any technology that surpasses 50% efficiency will never double its efficiency again.

We are already very close to 50% with 40-46% efficient solar cells. The installation cost, scaffolding cost, maintenance and the cell price start to dominate. Solar panel tilt can improve efficiency up to 30 percent (flat panel versus sun following 2-axis panel).

If you double efficiency you halve the size, roughly halving the installation, scaffolding, & etc. costs you mention.

Also those 40-46% efficient cells are expensive, requiring 3 different layers of semiconductors each tuned to a different band-gap. The panels we all buy for ~$1/watt are about 20% efficient, single-layer designs.

Thise multijunction cells are very exotic in their materials and manufacture. Carbon nanotubes sound easier in principle.

I this is actually true and feasible, this has immense implications. Imagine mirror powered airplanes.

>yield a 90% efficient solar cell

Yield a 90% efficient solace cell for a single wavelength

... per carbon nanotube, of which there are trillions, each tuned to a different wavelength.

Not true if you check the spectral absorbance chart they show...

edit: actually, I'm wrong, see below

You're right, I believe. I don't have time to investigate, but if the claim is 90% efficiency w.r.t. total radiation input, it's certainly wrong. Doing so would violate the 2nd law. For example, room temperature thermal radiation has significant emmitance in the radio range of the spectrum. Needless to say nobody can build radio antennas to turn this energy into work.

If it is 90% of the thermodynamic limit, then maybe. But then I think multijunction cells are already significantly close to the thermodynamic limit (much better than 22% claimed).

https://en.wikipedia.org/wiki/Carnot%27s_theorem_(thermodyna...

(obs1: substituting Tsun as Th and Tearth as Tc left as an exercise to the reader ;) )

(obs2: try changing Tc to CMB temperature! )

Those "significant" amounts of radio are an extremely small amount compared to the visible and near infrared output of the sun.

> substituting Tsun as Th and Tearth as Tc left as an exercise to the reader

Left to the reader because doing the math reveals that it's only a 6% difference and undermines your point? ;)

Oh thanks. I hadn't actually plugged in the numbers, shame on me :P

It seems the maximum efficiency in this case is ~94.8%, so the claim does seem plausible at least.

Well indeed. I assume that the tubes don't grow at a uniform height, which means wide band performance.

What I don't understand is how they pair up a rectifier with each antenna so that you don't cancel out the signal due to either path/frequency mismatch.

ELI5: What is a 'rectenna'

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

Radio waves jiggle the electrons back and forth in your chunk of metal. Stick in a diode so they can only go one way and you have power.

> hopes to use carbon nanotubes to...

> a process they hope will yield...

Yawn. I hope to get a new truck for Christmas- let me know when Santa Claus is real.

Nice in theory. Very difficult to get anywhere near that efficiency in practice. Best of luck to them, but I'm not holding my breath.

Reply on news.ycombinator.com