The Solar Sunflower | Hacker News Reader

The Solar Sunflower

arstechnica.com

79 pointsRifu46 comments

I take exception to the use of the word 'invention' here because that most likely implies that patents are at stake. Tech such as this has been tried on-and-off by various parties since the 80's, with many variations on geometry, working fluid, actuation and tracking.

Parabolic segments are a pretty expensive way to do this because it requires moving the collector around. Much simpler and easier to have a stationary reflector harness and actuate the individual mirrors instead, then you can keep the collector on the ground where it is much easier to tie it into something that makes use of the collected energy.

Agreed. The simpler solutions are cheaper and easier to maintain. There are just too many points of failure on a system such as this.

Especially if such reflector units could be mass produced and designed such that their mooring+maintenance requirements are absolutely minimal.

I built a (crappy) prototype to test this and it worked remarkably well:

http://pics.camarades.com/v/jacques/renewables/concentrator/

That is pretty cool, it looks remarkably similar to something one of the Google Energy projects was doing. I tried something a bit different which was to make an actual parabola but to build a mirror reflected the Sun into the same point all the time. (basically you take advantage of the fact that the position of the Sun in the sky for all time/seasons can be calculated for a particular latitude, and then build a reflector where it has a surface for each calculated point.) It works ok, and as a bonus you have something you can put in your lawn which is always annoying your neighbor :-)

The real magic in the sunflower though is that its a dual cycle system harvesting both thermal and electric energy from the system. A lot of "off grid" solar installs will have both PV and Thermal panels as burning electricity to create hot water is really wasteful.

> The real magic in the sunflower though is that its a dual cycle system harvesting both thermal and electric energy from the system.

I did the same with that setup above, the collector was water cooled and the hot water could be used. Taught me a ton about what you can and can not do to a normal (cheap) solar panel from a thermal point of view. That's a lot of suns to concentrate on a single panel and the effect on the life span of the panel was quite pronounced, even with sufficient cooling.

The big trick (at least, so I thought) was that I figured out a gearing that would allow you to drive that whole setup with one motor and a small solar panel.

Agreed. I saw a demonstration of a unit very similar to the one in the article, that had vacuum-focused thin film reflectors and a sodium-cooled sterling engine. That was in 1997.

This sort of attitude really bums me out. "Tried it!"

With the current rate of technological advance, a ton of stuff that's failed (and even a lot that's succeeded) should be tried again. Frequently some incremental but necessary technological improvement that didn't exist 10 years ago exists now and it might just change the game. Maybe that's happening here?

Perhaps incremental improvement doesn't deserve the word "invention" but it still deserves a look.

People who are trying things that already failed should mention why it failed previously and why it will not fail this time. At least journalists who write about such things should be held to that standard.

The 'invention' part of this setup is not their mirror configuration, which just looks pretty and possibly is easy for them to build and deploy in spaces where space is a problem. You have to admit that it looks good in that sunflower field. Their footprint looks like it's <2m^2, sort of like a billboard, that makes it very attractive to place wherever.

As it says in the article, the actual interesting part is how they generate the energy, apparently with a GaAs that's cooled with some sort of futuristic microfluidic water system. If that works well just the receptor can be a very profitable product, they can let others worry about mirror configurations and deploying in farm lands.

Maybe the application but tech wise not seeing much new, heat exchanger for electrical generation, satalite dish tracking of a signal (albiet light spectrum).

Maybe if the mirrors had prisms which would focus the light frequency for solat to the cells and would focus the thermal at another spot, that would be neat. Then could of had reflective panels that absorbed the heat and processed there and only reflected the light for solar convertion.

Now be nice if solar heating systems and panels were combined, could have a lens like tube that would absorb the heat but reflect the light towards a solar panel.

Though for me why solar panels do not have large lenses infron as a form of even plastic lens would or glass would work out cheaper than some panel tech and with that use less panels and use less expensive elements and be cheaper over say plastic/glass used to focus. Well that is something I do ponder.

Still hope it works out and if it don't then they will learn lots. But it is sort of just of taking the Archimedes death ray approach http://web.mit.edu/2.009/www/experiments/deathray/10_Archime...

So as you say it will be down to the conversion of light and thermal at one point. But for me that whilst useful (more so in space perhaps) is for me at least complicating a problem in which other approaches would seem easier to engineer (ie collect heat in the panels).

> I take exception to the use of the word 'invention' here because that most likely implies that patents are at stake.

Why shouldn't patents be at stake? Maybe they've come up with something new and non-obvious, or maybe they haven't... we can't exactly tell from this article though.

> Tech such as this has been tried on-and-off by various parties since the 80's, with many variations on geometry, working fluid, actuation and tracking.

And if all those folks didn't succeed, isn't that more reason to give Airlight a patent? Before some competitor realizes that Airlight has fixed the thing everyone else couldn't, then shamelessly copies them?

Patents should not be linked to commercial success, if that's what you mean by "didn't succeed".

Lot's of people try and make cancer cures, in theory only the people that succeed get useful patents.

There is no inventive step. Just some engineering. They should get patent on what exactly? Cooling semiconductors? Concentrating sunlight?

I wonder why they are bothering with gallium-arsenide panels when using a solar thermal setup would be a lot more cost efficient and probably more energy efficient as well. You simply run some kind of coolant through the collector, exchange the heat into water, and run a steam turbine from the resulting steam. You can even collect the water afterwords and condense it to get distilled water. There are many power plants that run on that principle, but it would be cool to see something that can be mounted in the back yard. And the materials necessary are very common and cheep, so the entire product could be much cheeper as well.

Well, you'd end up with something like this:

https://www.mtholyoke.edu/~wang30y/csp/ParabolicDish.html

GaAs panels have ~40% efficiency. It's not easy/dirt cheap (you need high temperatures, and materials to go with that) to get that much efficiency out of steam turbine (or Stirling engine). For example your typical nuclear power plant would have about 33% thermal efficiency. And with the panels you have less moving mechanical parts that need servicing. And the panels are silent vs. steam turbine. Lighter also. Pure solar thermal has been tried/used many times before.

On the other hand, their setup claims to melt iron, at 1518C. Which would give you a pretty good carnot efficiency if you'd use it directly to fire a steam turbine.

> GaAs is much more efficient at converting sunlight into electricity (38 percent in this case, versus about 20 percent for silicon).

Is that true... You get almost 2 times the power output with GaAs cells then conventional solar cells made from silicone?

Yes[0]. Apparently the records lie at 41% for GaAs and 27% for silicon[1] so they're being a little bit loose with the stats, 20% for silicon is consumer grade panel level.

0] https://en.wikipedia.org/wiki/Gallium_arsenide#Solar_cells_a...

1] http://www.pveducation.org/pvcdrom/appendices/solar-cell-eff...

The sub-title claims 80% efficiency. Where did they get that from? 80% is close to the theoretical maximum, so I think the figure is misleading and can discredit Arstechnica and more importantly an otherwise admirable endeavor.

That's possible because they count both photovoltaic and thermal conversion efficiency. I'm not entirely clear on whether they want to use the heat as heat or always for electricity generation. They claim GaAs solar cells get 40% efficiency, which I'm ready to take at face value. 60% of the energy is then converted to heat. For a total of 80% conversion efficiency we'd need only 66% efficiency for the thermal-to-electricity conversion, which is well within theoretical reach, given they also claim >1000C achieved temperatures.

They can't convert the heat to electricity because they can't run the solar cells that hot - max operating temperature is 105C, and I'm not sure how great the efficiency and lifespan would be at that temperature. The 80% efficiency assumes they use the heat for heating, which might be a bit difficult in practice given the times of year and locations where this could actually be operated.

Oh. Well, if that's so, I'm not sure why they've chosen to use photovoltaics at all, because solar thermal should give you a much better cost-performance ratio, and still give you plenty waste heat to use as heat.

While reading the article, I was wondering why combined PV/water heating conventional panels aren't taking off on rooftop installations. Looks like something that would make a lot of sense.

It's a good question. The panels are available, for example these: http://www.solimpeks.com/pv-t-hybrid-solar-collectors/

My guess is they're not as popular because the overall installation complexity is much higher: you have to run the piping, and I'm not even sure exactly what you do with the hot water that comes out. There won't be room for it in your existing hot water heater, so you'll need a second tank, and where do you put it? How big does it need to be? And what happens when it fills up? Are the panels cooled only with fresh cold water, or do you reuse the water in the second tank once it's full? And if the latter, what happens when that water gets too hot to use? I guess you'd start letting some of it evaporate -- that'll be the fastest way to get rid of the excess heat, but may be undesirable in a water-poor area; and once you start evaporating water, you have to worry about scale formation.

Probably somebody knows the answers to all these questions, but they're not obvious to me. For new construction, they're probably not that hard to solve, but for a retrofit on an existing house -- which is the vast majority of where the panels are going -- the complexity seems prohibitive.

Somebody please correct me if I'm mistaken!

Yes I'm sure there is some good reason, but considering that solar for water heating - without PV - has been sold for ages already, there should be a market. Let's hope somebody with a real-world understanding of the problem will answer, I'm really curious :)

Swimming pools - in areas like the NC Outer Banks - the systems give you an extra month or so of swimming on each end of the season.

The heating water part is really common here in Germany and not that big of a deal, and not just for new constructions. Basic heat storage/exchange parts often easily fit in existing rooms or even in the space the old heating system took.

You only see either solar cells or heat collectors though.

Why there are no/not many combined setups, no clue. Maybe the efficiency goes down to much and limits the range of regions where it is viable? Maybe there are regulatory issues with water + power systems so close together (driving up installation and maintenance prices).

> Maybe the efficiency goes down to much and limits the range of regions where it is viable?

Actually, efficiency for PV would go up (the colder the cell, the higher the efficiency).

> Maybe there are regulatory issues with water + power systems so close together (driving up installation and maintenance prices).

Maybe, that would be interesting to learn :)

I don't know if cells would be colder, since the heat collection parts works at liquid temperatures of 80°C or if they don't use water directly even over 100°C. The heat collecting system wants to be as hot as possible, the solar cells as cold as possible. Also, the cells decrease efficiency of the heat collection part.

If you get the heat from the cells, they need to become colder for sure.

...I'm not even sure exactly what you do with the hot water that comes out. There won't be room for it in your existing hot water heater, so you'll need a second tank, and where do you put it?

You recirculate water from your hot water tank through a heat exchanger that receives heat from the solar panels via a closed fluid loop. If the heat exchange would cool rather than heat your water, a thermostat turns the pump off.

Most of the new installations are linked to energy leasing vehicles. Water doesn't work there. Also, most folks use gas or propane to heat water, and gas prices are very low.

People spend a lot money in heating water. My folks did a hot solar hot water system in the 90s, and I think that the payback period was something like 3 years.

The system exchanges a fluid (I think glycol, but don't quote me) into a coil in front of the water heater water supply. So instead of dumping cold water in the tank, you're feeding it warm water. With a tankless system, it's even better.

I don't have a link, but combined Solar Electric / Hot Water panels are relatively common Australia. My neighbor has them on his new house. I'll go ask him for details.

Thanks, that's interesting to know. Never saw any promotion about them here in Italy.

Ack, those circles waste so much space. They should put a in an array of hexagonal reflectors instead.

Wouldn't that give you a flat surface in total?

I would venture to guess that it's a manufacturing issue. The mirrors are parabolic, and there probably hasn't been much call for hexagonal parabolic mirrors before now. Perhaps they'll develop that infrastructure if it takes off.

Another step towards _Ringworld_!

The gigantic PS power plants in Seville are the same thing, using a water tower instead photovoltaics.

They are quite spectacular: https://en.wikipedia.org/wiki/PS10_solar_power_plant

They're not, as the whole point of the "Sunflower" is the integrated GaAs PV collector. The Seville CSP plant is a conventional concentrated solar (thermal) power plant: energy is generated via heat exchange and steam production using turbines, not from photovoltaic effects.

A bit of background from some earlier articles I've seen added to this;

The concept is based on IBM technology aimed mostly at thermal management in coimputer servers, but with applications to solar power. See e.g.,: https://redd.it/2afct9

Traditional PV has the problem that elecrical generation is degraded at higher temperatures. This is especially the case where sunlight is concentrated on the collector. That's why most PV are flat panels rather than arrays of (cheap) mirrors + panels. Another factor is that flat panels don't need to track the sun (you point them generally south and angled by your latitude), hugely simplifying installations.

The "5,000 suns" bit here basically means that a total reflective surface of 5,000x the PV area is used. This also means that heat is approximately 5,000x greater. An advantage is that you need far less of the expensive PV collector. A problem such designs is encountering is that PV really isn't that expensive anymore. Oh well.... Of problems to have, this is a relatively good one.

Getting both electricity and heat from the same installation, if you've got something to do with both, can be useful. Though it often isn't. As the images accompanying the article show, the "sunflower" design ... isn't exactly unobtrusive. Where panels can be designed directly into new or existing construction, studding your walls and roof with heleostat parabolic dishes is ... more challenging.

There are also integrated solar PV + thermal systems. As others have noted they increase complexity and installation costs, though if you can stomach those the benefits in terms of overall fuel and mains savings are substantial.

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

A few further points:

● The real key in solar isn't efficiency but cost. Single-layer solar is limited to about 37% efficiency because physics, the peak theoretical efficiency is about 85%, at high cost, again, because of physics. You're starting from a ~1kW/m² flux, with reductions from there for panel efficiency, spacing factors, capacity factor, inverter (direct DC use would be a 10% gain), and net storage costs. The 85% efficiency gains of the Sunflower would all but certainly be eaten up through installation, maintenance, and manufacture costs.

● Solar and thermal can be combined, and with sufficient thermal mass, gains are fairly substantial.

● Total collector area matters. A small number of high-effieciency collectors ... collects a small amount of energy. A large, cheap area of modestly lower-efficiency collectors beats the efficiency gains. It's usually possible to increase collector area 2-10x. It's not possible to increase collector efficiency 10x starting at a 20% baseline.

For some massively impressive examples of what can be done with net-zero construction in Fairbanks, Alaska, see Thorsten Chlupp's Reina LLP:

http://www.reina-llc.com/

I particularly recommend the video series. http://www.reina-llc.com/resources/videos1/

The following two videos are long (2+ hours) but incredibly detailed. Particularly impressive are Chlupp's use of thermal mass (both static insulation and a central thermal-stratification 5,000 gallon tank), and the consequent challenges (air exchange, heat exchange on all interfaces, including freshwater and sewerage), and most critically: moisture control -- condensation across a 100+ degree thermal gradient is a massive challenge.

If you're looking for solid construction details, these are an excellent resource.

Path to Net Zero Energy Series -- Alaska's first Net Zero Homes https://www.youtube.com/watch?v=AtHkvpRI6fc

Alaska's First Net Zero Energy Homes Performance Update https://www.youtube.com/watch?v=Xen_VWyDezY

2nd video is an update on the home described in the first. Goal achieved, except for plug-in Prius accounting.

Is the "5,000 suns" part complete marketing BS? The description says that they compress the energy of one sun into 1/5000 the area, to reduce the expense of photovoltaic material. But the process isn't 5000x more efficient at obtaining energy from the sun.

At best, it's "the power of 5000 solar panels", which is a much less exciting headline.

We've taken that bit out of the title here.

Reply on news.ycombinator.com