This is a questionable way to present what's an excellent project and hopefully soon to be commercialized technology. The big deal here is it's a presumably installation ready application of EDR for desalination instead of RO which most systems use. This is a big deal because the membranes use electricity instead of pressure as the filter, which means everything can run at low, normal plumbing, pressures instead of the crazy high pressure RO stuff. For seawater it's borderline whether or not it will match RO for performance, but for lower salinity groundwater and industrial wastewater, it should be significantly higher performance for the same power as well as lower maintenance and capex.
The no batteries thing is basically irrelevant to the innovation, and in fact Genius Water already offers no battery RO systems, also with questionable benefit (as well as being difficult to work with).
I run a solar and water focused EPC in East Africa and will hopefully be working with these guys in the future when they're off the ground with a commercial system. The potential is extremely high, particularly if the maintenance overhead and operational complexity can come down in practice.
ricksunny 5 days ago [-]
It sounds lime an MPPT on the supply side, with an ideal load-point tracking on the demand side. My understanding is that there are controllers (including for solar water pumping in the East Africa market) that pursue this. The concept applies more generally to systems where the load presented is configurable by system plant parameters, such as flow-rate & height.
hedora 5 days ago [-]
Yeah; the solar part is really questionable. In an installation without batteries, they’d need an additional large tank to store excess daytime output.
Without such a tank, they’d need to somehow power the thing at night, which means a big battery, just like RO.
Also, the article suggests the power input needs to be steady and they use a computer to run it at higher rates when the battery would be charging.
Assuming there is a small battery or power grid (as both systems require), you could oversize an RO system and then change its duty cycle to keep the batteries at (say) 80% to prevent the solar production from curtailing. Round-tripping electricity through our home battery loses about 20%.
So, the “advantage” boils down to two questions that the article doesn’t answer: (1) what are the relative energy efficiencies of this system (in theory) vs RO? If the new system is 20% worse, RO wins, regardless of this optimization (2) what is the relative equipment cost vs. max throughput? (Since both setups assume oversizing to get better solar utilization).
I’d also like to know if the new system requires plastic, since the RO membrane probably leaches all sorts of nasties into its output.
I do like the fact that they are focusing on brackish water. We have this problem even in the coastal US (in the form of water softener output), and I’m sure they could sell a premium alternative to RO as a way to get scaling advantages on the manufacturing of the equipment.
cryptonector 5 days ago [-]
Tanks are fairly cheap, and you'll need one anyways. But yeah, the solar angle is not what's interesting here. It's the electrodialysis.
elzbardico 5 days ago [-]
Unlke some comments are implying, this is not a solar distiller with "additional steps". It still uses far less energy than distillation as it doesn't involve phase changes.
It uses Electrodialysis, which is a mass separation process in which electrically charged membranes and an electrical potential difference are used to separate ionic species from an aqueous solution and other uncharged components.
SoftTalker 5 days ago [-]
Can't a lot of the phase change energy in distillation be recovered by using incoming water to cool/condense the distilled water vapor?
elzbardico 5 days ago [-]
It would improve it a bit, but not enought to compete in terms of energy usage with dialysis or reverse-osmosis.
Reverse-osmosis is absurdly efficient compared to distillation: a single 1X1 meter square solar panel can potentially generate 200 liters of fresh water per day.
bastloing 5 days ago [-]
That's great news! Now if they can solve the same problem with sea water, California, Arizona and Nevada can reduce their reliance on the Colorado river and grow more crops. It is only a matter of time before it's solved. Great work, MIT!
ttyprintk 5 days ago [-]
It’s a great application, but electrodialysis on seawater takes more power—-so much that distillation is competitive. The use-case chosen is remote freshwater wells which suffer from naturally-occurring arsenic. I can only think of a few others which can’t have heavy batteries.
This isn't really accurate, they're targeting industrial wastewater yes but they are working with and have tested brackish water up to several thousand TDS. They had a working EDR system for drinking water installed in Gaza until relatively recently and several in India as well. I'm also skeptical they can make it work with seawater, but it absolutely works with undrinkable brackish water in many other cases too.
5 days ago [-]
dylan604 5 days ago [-]
Getting water to heat/boil is much less impressive than coming up with a solve for the left over salt/minerals. Solve that, then I'll join in the "Great work"
SoftTalker 5 days ago [-]
> the left over salt/minerals
There is a commercial market for salt -- and for stuff like treating roads in the winter it doesn't have to be very clean.
Otherwise, disolve it into the local waste water stream and discharge it back into the ocean.
dylan604 5 days ago [-]
If this was the case, then why is the briny residue left after desalination always the thing that gets pointed back to being a big negative of desalination?
Either it's not as big of deal as people suggest, you are wildly underplaying it, or somewhere in between. I've never felt that the argument against being the cost to heat the water was a strong one since salt water pretty much means a coastline which tends to have steady wind and sun. The biggest hang up has typically been putting that brine back into the ocean.
fred_is_fred 5 days ago [-]
Even if you could do this you'd have to pump the water back uphill to NV and AZ.
itscrush 5 days ago [-]
No, start with modifying the Colorado River Compact and other underlying agreements to allow more upstream retention than is currently allotted.
aidenn0 5 days ago [-]
Could you save on pumping energy by sending the water to underground aquifers rather than the surface?
dylan604 5 days ago [-]
Just install some additional solar powered pumps along the way
fred_is_fred 5 days ago [-]
The amount of energy needed to pump enough water for ag uphill is insane. Well beyond "just" throwing some solar panels out there. If it was that easy we'd pump Mississippi water into west Texas (which there was a plan to do in the 60s with nuke plants, but I cannot find the name right now).
dylan604 5 days ago [-]
i left off the /s as to me anytime someone starts a comment off with "just ____" is usually a farcical idea. like just remove the salt from water and boom, done.
fred_is_fred 5 days ago [-]
Hah! It can be a very HN comment to "just" something especially around physical engineering.
great article but it tries to (ahem) separate drinking water from other uses, which doesn't seem practical:
- in the poorest places, they can't afford desal.
- in non-poorest places, most water is delivered by unified piping systems due to cost and labor efficiency. Schlepping water in bottles and buckets is nuts, though I can see it turning into the next weird fad in exercise or robotics.
JCharante 5 days ago [-]
This seems.. simple?
netbioserror 5 days ago [-]
While it's an odd example for this place, I can bring up self-loading firearms (semi-automatic or automatic in today's terms) as a demonstration. Modern self-loading firearms are VASTLY simpler than the early attempts a century ago. They're an excellent example of engineering evolving under economic pressures.
Late 19th and early 20th century attempts at self-loading firearms were often ridiculous in their concepts; huge component counts, lots of tiny mechanisms, strange attempts at extracting recoil and gas energy, everything under the sun. The mechanisms engineers were crafting in literal garage workshops are stunning in their variety and staggering in their watch-like complexity. Some were genuine works of art.
Then the M1 Garand, the SVT-40, and afterwards the AK (under the economic pressures of WW2) demonstrated how much room there was to simplify and give various components double duties. Now, most modern automatic weapons derive from those designs, and the improvements since have been in the materials engineering: Stronger, lighter, thinner, and generally reducing the amount of steel to the minimum necessary.
MisterTea 5 days ago [-]
The AK copied the STG-44 Sturmgewehr (literally "assault rifle, this is where the design and name comes from) which was revolutionary in design and abilities. Prior to the assault rifle solders weapons were either accurate long range rifles with high power cartridges or close range inaccurate sub-machine guns firing low power pistol cartridges. Military researchers realized that most solders were average people and could not make full use of the high power and accuracy. The solution was an intermediate cartridge that combines the longer range and accuracy of the rifle cartridge with the smaller profile and lower recoil of a sub machine gun. Now you have a weapon that can hit at a distance or go auto and fight close quarters. Huge advancement and advantage for the solders wielding such weapons. Kalashnikov was directly inspired by these abilities and developed the AK in response. Just about every modern "Assault rifle" is descended from the STG-44, not the AK.
netbioserror 4 days ago [-]
The AK arguably took more influence from the M1 Garand, given its rotating bolt, locking lug arrangement, and long-stroke piston. The STG-44 definitely proved the effectiveness of an intermediate round to lay the groundwork for the form-factor.
MisterTea 4 days ago [-]
Indeed. I should have mentioned that it was in fact inspired by both the M1 Garand and STG-44.
netbioserror 4 days ago [-]
Apologies the STG-44 is long-stroke, for some reason I mixed it up with its successors.
aidenn0 5 days ago [-]
Off topic, but it seems like self-loading pistols took a weird detour; at least for cartridges too powerful for blowback operation. There are all sorts of weird delayed-blowback systems that were popular between WW2 and 1980-ish, and now 9mm and larger seems to almost exclusively use a 1911-style short-recoil system.
netbioserror 5 days ago [-]
It's simple, reliable, and quite necessary. Pistol chamberings feature heavy bullets in straight-walled, short cases. Blowback bolts are always extremely heavy to compensate for those attributes. Beretta and FN are famous for resisting Browning short-recoil for alternatives like rotating barrels and locking blocks. But they pay for those tradeoffs: Heat buildup, wider slides/frames, extra complexity, and more. Browning short recoil is the best of all worlds. Replacing rotating links with simple cam cuts sealed the deal.
aidenn0 5 days ago [-]
I actually think the gas-delayed blowback in the HK P7 hits "simple and reliable" as well, but it has the huge downside of putting very hot gasses very close to where you handle the gun.
WJW 5 days ago [-]
Many useful things are simple. Not all of them exist yet.
jvanderbot 5 days ago [-]
Yeah if the main insight is "you can run electric dialysis desalination on variable input power" they sure did a lot of dressing it up.
devilbunny 5 days ago [-]
Desalination that can start and stop, increase or decrease activity, without messing anything up is the secret sauce here.
Not going to do that with reverse osmosis systems.
That said, with merely brackish input water, I'm wondering how many problems this really solves. Drinking water, sure, but you have to get rid of the concentrated brine at the end and it's still groundwater that can be overdrawn.
However, if v 2.0 can effectively desalinate ocean water, it would be huge for islands and coastal areas.
marcosdumay 5 days ago [-]
The main insight is "we've spent the last years tuning a dialysis controller to work well on variable input power".
I imagine the paper has the actual parameters, so you can build upon their work.
> “The majority of the population actually lives far enough from the coast, that seawater desalination could never reach them. They consequently rely heavily on groundwater, especially in remote, low-income regions. And unfortunately, this groundwater is becoming more and more saline due to climate change,” says Jonathan Bessette, MIT PhD student in mechanical engineering. “This technology could bring sustainable, affordable clean water to underreached places around the world.”
Uh, that's just going to increase the rate of acquifer depletion.
surajrmal 5 days ago [-]
Reminds me of cloud based batch jobs. We must have many more opportunistic workloads similar to this one.
aaron695 5 days ago [-]
[dead]
lyragapad2009 3 days ago [-]
[dead]
voidUpdate 5 days ago [-]
This sounds a lot like the concept of a solar powered distiller... As in, heating a container of water with the sun, evaporating the water and then cooling it down to convert it into fresh water...
marcosdumay 5 days ago [-]
It's not. It's an electrodialysis desalinator. I have no idea what in the article gave you the idea it was thermal.
yogurtboy 5 days ago [-]
I think their thought process was "It uses the sun to desalinate water, so it must be the same"
black6 5 days ago [-]
But with extra steps and points of failure.
EricE 5 days ago [-]
I think you are missing the key point: it takes a fraction of the energy vs. the solar desalinization you are referring to.
Rendered at 18:20:30 GMT+0000 (Coordinated Universal Time) with Vercel.
The no batteries thing is basically irrelevant to the innovation, and in fact Genius Water already offers no battery RO systems, also with questionable benefit (as well as being difficult to work with).
I run a solar and water focused EPC in East Africa and will hopefully be working with these guys in the future when they're off the ground with a commercial system. The potential is extremely high, particularly if the maintenance overhead and operational complexity can come down in practice.
Without such a tank, they’d need to somehow power the thing at night, which means a big battery, just like RO.
Also, the article suggests the power input needs to be steady and they use a computer to run it at higher rates when the battery would be charging.
Assuming there is a small battery or power grid (as both systems require), you could oversize an RO system and then change its duty cycle to keep the batteries at (say) 80% to prevent the solar production from curtailing. Round-tripping electricity through our home battery loses about 20%.
So, the “advantage” boils down to two questions that the article doesn’t answer: (1) what are the relative energy efficiencies of this system (in theory) vs RO? If the new system is 20% worse, RO wins, regardless of this optimization (2) what is the relative equipment cost vs. max throughput? (Since both setups assume oversizing to get better solar utilization).
I’d also like to know if the new system requires plastic, since the RO membrane probably leaches all sorts of nasties into its output.
I do like the fact that they are focusing on brackish water. We have this problem even in the coastal US (in the form of water softener output), and I’m sure they could sell a premium alternative to RO as a way to get scaling advantages on the manufacturing of the equipment.
It uses Electrodialysis, which is a mass separation process in which electrically charged membranes and an electrical potential difference are used to separate ionic species from an aqueous solution and other uncharged components.
Reverse-osmosis is absurdly efficient compared to distillation: a single 1X1 meter square solar panel can potentially generate 200 liters of fresh water per day.
There is a commercial market for salt -- and for stuff like treating roads in the winter it doesn't have to be very clean.
Otherwise, disolve it into the local waste water stream and discharge it back into the ocean.
Either it's not as big of deal as people suggest, you are wildly underplaying it, or somewhere in between. I've never felt that the argument against being the cost to heat the water was a strong one since salt water pretty much means a coastline which tends to have steady wind and sun. The biggest hang up has typically been putting that brine back into the ocean.
- in the poorest places, they can't afford desal. - in non-poorest places, most water is delivered by unified piping systems due to cost and labor efficiency. Schlepping water in bottles and buckets is nuts, though I can see it turning into the next weird fad in exercise or robotics.
Late 19th and early 20th century attempts at self-loading firearms were often ridiculous in their concepts; huge component counts, lots of tiny mechanisms, strange attempts at extracting recoil and gas energy, everything under the sun. The mechanisms engineers were crafting in literal garage workshops are stunning in their variety and staggering in their watch-like complexity. Some were genuine works of art.
Then the M1 Garand, the SVT-40, and afterwards the AK (under the economic pressures of WW2) demonstrated how much room there was to simplify and give various components double duties. Now, most modern automatic weapons derive from those designs, and the improvements since have been in the materials engineering: Stronger, lighter, thinner, and generally reducing the amount of steel to the minimum necessary.
Not going to do that with reverse osmosis systems.
That said, with merely brackish input water, I'm wondering how many problems this really solves. Drinking water, sure, but you have to get rid of the concentrated brine at the end and it's still groundwater that can be overdrawn.
However, if v 2.0 can effectively desalinate ocean water, it would be huge for islands and coastal areas.
I imagine the paper has the actual parameters, so you can build upon their work.
Uh, that's just going to increase the rate of acquifer depletion.