By Brian Sheng, Co-Founder and CEO, Aquaria. Trained as an engineer at Princeton, where he studied water technology; co-inventor on Aquaria's atmospheric water generation patent application (USPTO App. No. 64/009,115).
TL;DR --Yes, an atmospheric water generator runs on solar, off-grid or on it. Three things decide whether it works at your property: whether your worst month gives the machine enough warmth and humidity to produce well (below roughly 13°C and 25% relative humidity, output falls off hard), whether you can afford the array, and whether you buffer with a battery or a water tank. Buffer with the tank where you can. Water storage costs about a dollar a gallon before installation, it holds for a season without losing anything, and a battery cannot do that at any price. You still need some battery for compressor startup, controls, pumps, and cloudy stretches, so the honest claim is “no large battery,” not “no battery.” And if no electricity reaches the device at all, you are looking at a passive panel making liters per day, not a household supply. Run the four-step site check below before you price anything.
Step 1: Will It Work at My Property?
Start with what a machine actually makes where you live, not with a humidity chart. Below is modeled output for one Aquaria machine across a range of US climates: the yearly average, and the weakest month, which is the number that decides your system. Find the row that looks like your climate.
Modeled from NOAA 1991-2020 climate normals, one machine, before any allowance for downtime. Twenty markets are in the full dataset.
How to read your row. Compare the weakest month against what you actually want per day. If a family wants around 40 gallons a day, Miami, Honolulu, and Tampa clear that in every month of the year, so a modest tank is enough. Houston and New Orleans clear it easily for most of the year but dip below it in January, so you either store water ahead of winter or lean on a second source for a few weeks. Atlanta and Raleigh produce well for eight months and stop in the deep winter, which means a real tank or a real backup. Phoenix, Denver, and Las Vegas are warm-season or supplemental sites, and the honest read is that air is the limiting resource there, not your equipment.
Notice the gap between the first two columns. That gap, not the yearly average, is what you are designing around, and it is the reason a system sized on an annual number disappoints its owner in February.
If your city is not listed, the pattern that matters is warm and humid versus cool or dry. Coastal and Gulf climates behave like the top of the table, inland arid and cold climates like the bottom, and everything else falls between. Send us your town and we will run your twelve months.
Why Winter Hits So Hard
Cool, dry air simply holds less water, and below roughly 13°C and 25% relative humidity a machine's output falls off steeply. It keeps running, it just makes less, and plenty of owners run machines in those conditions anyway because some water beats no water. What changes is how much of the day is productive.
Miami stays productive around the clock in all twelve months, which is why its weakest month is still 186 gallons a day. Houston does the same from April through October, then drops to about ten productive hours a day in January, which is where its 39-gallon January comes from. Atlanta runs a full day in July and a few hours in December. That is the whole story behind the first two columns of the table above.
Step 2: Three Ways to Build It, and What Each One Buys
There are three ways to run a water-from-air machine on solar, and choosing between them is the decision this article exists to help with. The energy industry calls this choice dispatch, which just means deciding when a machine runs. The three setups differ in when the machine runs and therefore in whether a battery or a water tank carries your household through the hours it is off. All three work. They are not better and worse, they are different distributions of the same money.
The first setup works because the machine is schedulable from the Aquaria app on time of day, humidity, and temperature, so running it in your solar window is something you set rather than something the weather decides for you. That single capability is what makes the buffer a real choice instead of a theoretical one.
The third deserves more credit than it usually gets. Every off-grid property already has an answer to “what happens when the resource disappears,” and for most of them it is a generator. Sizing the array and the tank for the single worst week of the year is expensive; covering that week with fuel you already store is often not. How large each piece has to be, and what the machine and install cost, are covered separately.
What Happens to Excess Solar When the Batteries Are Full?
In most off-grid systems, it is thrown away, and an atmospheric water generator is unusually good at catching it. Once the battery is charged and the household loads are met, extra production has nowhere to go, so the system simply turns it away. Installers call this curtailment, and on a well-sized off-grid array it happens most sunny days of the year. That surplus is not small, because the array was sized for the worst month and most months are not the worst month.
Take Houston. A rooftop array there produces about 2.6 kWh per day per kilowatt installed in its lowest month and about 3.8 on the annual average (NREL PVWatts, via our sizing piece). Size the array to the lowest month, as you should, and on an ordinary day it makes about 46% more than the water needs. That margin is not waste in the usual sense, because you had to buy it to survive December. It is simply sitting there most of the year.
Two things make water an unusually good home for that surplus. It is a load whose output you actually wanted, unlike a resistive dump load that turns surplus into waste heat. And the surplus season coincides with the machine's best season: in our capacity dataset Houston sits above the derating point 24 hours a day from April through October, which is exactly when the array is running its largest surplus. The seasonal timing works in the owner's favour, which is the opposite of what the category usually assumes.
This is where the Aquaria app matters; it runs the machine on time of day, humidity, and temperature, so you can point it at the middle of the day and tell it to work when the air is actually worth working, which in most climates is the same stretch of months your array is running its biggest surplus. Set it once and the surplus becomes stored water instead of nothing.
The machine follows the clock and the air, not your inverter. It does not read your battery state of charge and decide to run on its own, so this is a setting you make rather than something that happens automatically. In practice, midday through the good months is a close enough proxy, and it costs you one afternoon of setup.
Step 3: Do I Need a Big Battery?
Not necessarily, and a battery sized to run the machine around the clock is usually a wasted purchase. An atmospheric water generator turns electricity into a product you can stockpile, so the water tank can absorb the hours the battery would otherwise have to cover. Running mostly on the solar window and banking water is a legitimate design, and it comes down to five things holding for your site.
Two of them you already know: whether your tank can carry the evening and overnight draw, and what your backup is for a long stretch of bad production. The other three are about the machine and your power system fitting each other, starting the compressor, running on the power your array actually delivers, and producing well at your daytime conditions. Those depend on your specific site, array, and climate together, so they are worth settling with an Aquaria Water Advisor before you buy any machine. Bring your location and your power setup, and leave with a yes, a no, or exactly what to change.
If the fit is there, the battery shrinks to what startup and controls require, a much smaller purchase than a bank sized for 24-hour runtime. If it is not, size the battery conventionally and treat the tank as a way to reduce it rather than replace it.
What Still Needs a Battery, No Matter How Large the Tank
Five loads need electricity that no tank can supply: compressor startup surge, controls and sensors, the delivery pump, bridging cloud cover, and any deliberate nighttime running.
Startup surge is a power requirement rather than an energy requirement, so a larger tank does nothing for it. Controls need continuous power or the machine cannot sense conditions and restart on its own. The delivery pump draws when the household draws water, disproportionately in the evening after the array has stopped, so a tank shifts when the water was made but not when the pump runs.
The crisp version: a tank replaces kilowatt-hours. It does not replace kilowatts. Kilowatt-hours are fungible between the two media; kilowatts are not. How large the battery and inverter actually have to be is a sizing question rather than an architecture one, and it is covered separately.
Step 4: Bigger Tank or Bigger Battery? Where the Money Should Go
A solar water-from-air system is really a choice of what you store: electricity in a battery, or finished water in a tank. The machine is unusual among household loads because its output keeps, and that changes what counts as a good time to run it. Electricity in a battery has to be worth storing and it quietly loses value while it sits. Water in a tank does not. So the two media are not competing on price for the same job, they are doing different jobs, and the one that matters most off-grid is the one a battery cannot do at all.
With a battery, electricity you store has to be worth storing, and it leaks value while it sits. With a tank, any hour the machine runs is an hour of permanent gain. That includes mediocre hours. If your array is producing power you would otherwise throw away, running in air that is merely acceptable still banks real water, and a slow hour of surplus production costs you nothing but adds to the tank. Over a season that arithmetic is generous rather than strict: you are not waiting for perfect conditions, you are collecting whatever the air gives you on power that had no other use.
The clearest way to see the difference is not price. It is duration.
A battery does not do seasonal storage at any price. A tank does. Our sizing model puts Atlanta's seasonal buffer at 3,929 gallons for a single machine at a 42 gallon per day target, modeled from NOAA climate normals. Making that water took about 3,575 kWh of electricity at 0.91 kWh per gallon, and all of it is sitting in the tank, banked as finished water. For scale, a Tesla Powerwall 3 stores 13.5 kWh, so the tank is holding the energy equivalent of roughly 264 of them. Nobody sells residential electrical storage at that scale, and it would not help if they did: batteries self-discharge, and they age on the calendar whether you cycle them or not. Water in a sealed tank does neither. It is still there in January.
That is the argument, and it is a category difference rather than a price difference. The two media are not competing on cost for the same job. They are doing different jobs.
Where each one is genuinely the right answer. Buy battery for the hours: startup surge, controls, pumps, bridging a cloudy stretch, and running after dark if you want to. Buy tank for the weeks and months: the shoulder season, the weak month, and the surplus you would otherwise curtail. Water storage runs about a dollar a gallon before installation, which is Aquaria's own install experience, so the seasonal buffer is the cheaper end of the system rather than the expensive end.
The last row of that table is the honest limit, and it matters more than anything above it. A tank substitutes for battery capacity, not for array capacity. A gallon of tank helps only if the array and the air can produce the surplus gallon that fills it. Undersize the array and no tank closes the gap. Sit far below the derating point for a whole month and no tank closes it either, because there is no surplus to store.
Water keeps. Electrons do not.
That is the last time this piece uses the tank-as-battery framing, and no other piece in this cluster leads with it.
What the Machine Draws, for the Reader Doing Their Own Math
Aquaria's Hydropack X uses 0.91 kWh of electricity per gallon at its rated 30°C and 80% relative humidity, on 2026 shipping hardware.
In cooler or drier air the figure rises, because the machine runs longer for less water: directionally about 1.1 kWh per gallon in Atlanta, 2.3 in Seattle, 2.7 in Phoenix, 3.7 in Denver, and 8 in Las Vegas, derived from Aquaria's published 2026 per-state operating-cost model and stated directionally. An illustrative 42 gallon per day target in warm humid air is about 38 kWh per day. The average US residential customer used 865 kWh per month in 2024, about 28 kWh per day (EIA, 2024), so 42 gallons a day from air is roughly 1.3 times an average American home's total electricity use.
A module is rated at 1,000 liters per day (264 gallons) at 30°C and 80% relative humidity. What it draws moment to moment, and what your inverter and battery therefore need to be, depends on how the machine runs at your conditions, so it is a sizing conversation rather than a number to lift off a page. Warm humid markets land at the same 0.91 kWh per gallon on an annual-average basis as at the rated point, which is a coincidence of the published data rather than a restatement.
Which Constraint Are You Actually Under?
One thing before the constraints: you do not have to be off-grid for any of this to apply. Most people with solar panels and a battery are still on the grid, and everything in this article about running the machine on your own power, banking water in a tank, and putting surplus solar to work applies to a grid-tied house too. The difference is that your backstop is the utility instead of a generator, which makes the decision easier, not harder. And if you already own panels and a battery, adding water closes a loop neither one closes alone: energy from the sun, water from the air, both produced and stored on your own property. That is optionality and real independence rather than a backup plan.
“Solar or no electricity” compresses three different constraints, and they sit on one axis: how much electrical infrastructure you are willing to have. No grid means solar, battery, generator, or any hybrid is allowed, and only the utility connection is missing. That is the case everything above addresses, and it is what nearly every off-grid buyer means. No battery means the machine must follow available solar directly or run on a very small buffer, which is Step 3. No electrical input means no electricity reaches the device at all, so the process has to be passive.
One correction, because it is the most common error in this conversation. “No electricity” does not mean “no energy.” A passive device still runs on an energy gradient: sunlight as heat, radiative cooling to the night sky, or air movement. Nothing pulls water out of air for free, and the absence of a wire is a choice of energy source rather than an exemption from thermodynamics.
Passive and No-Electricity Options, Honestly
For a site with no power budget at all and a drinking-water need for one or two people, a passive sorbent hydropanel is a working answer and an active machine is not. That is a real niche, we do not serve it, and a competitor genuinely wins there. Hydropanels need no electricity, no inverter, and no battery, because sunlight supplies heat rather than power and there is no compressor to start. Radiative condensers and fog nets occupy similar territory where conditions suit them.
Now the boundary, with the same rigor. Global Water Intelligence, a market-intelligence firm covering the sector, has publicly stated there is no place on earth where a hydropanel is the most cost-effective source of water, and described the product as making very little water for its size and price. SOURCE, the largest hydropanel maker, publishes 2 to 5 liters per panel per day and recommends a two-panel array as its standard residential configuration, so roughly 4 to 10 liters a day (SOURCE published specifications, via World Water Reserve, 2026). Each panel is about 4 by 8 feet and 340 pounds, which makes this a roof-scale decision rather than an appliance purchase. Set that against a published benchmark: WHO guidance allocates 5 to 7 liters per person per day for drinking and food preparation, so four people need 20 to 28 liters (WHO/SEARO). At the bottom of its own published range a two-panel array covers under a fifth of that; at the top, roughly half. Drinking water for one or two people is the honest envelope.
Two related questions we get and should answer plainly. A dehumidifier is not a shortcut. It makes condensate, not drinking water: no air-intake filtration, no disinfection, no food-grade wetted path, no mineralization. The condensation physics overlap; the safety engineering is the entire difference. And this piece is about buildings, not vehicles. RV, van, and tiny-home water from air is a different class of machine at a different volume, and whole-home figures here do not transfer.
So What Should You Actually Build
Pick which uses get air water, check your worst month, size the array to that month, buy the smallest battery the hardware permits, and the largest tank the site permits. Which uses get air water is its own decision, and worth making one use at a time rather than all at once: drinking and cooking first, then washing, then the rest of the house, then anything outdoors. We cover that decision separately.
Every step after that first choice is arithmetic. What this article decided is which medium to bias toward, and what to settle before you sign.
Bias toward the tank until a load forces you into the battery. Then buy exactly that much battery.
Where to Go From Here
If you want to see what your own site would produce, the table in Step 1 covers a range of climates, and we can run your twelve months for any town. If you are pricing a system, our cost-per-gallon breakdown is the honest version, including where water from air loses.
And if you have a site in mind, solar on the roof or not yet, talk to an Aquaria Water Advisor. Bring your location, your rough daily water goal, and what power you have or plan. You will get what the machine would actually do there, including a no if it is not the right fit. Talk to an Aquaria Water Advisor.
Frequently Asked Questions
Can you run an atmospheric water generator on solar?
Yes, and off-grid it is a normal configuration. Aquaria's HydroPack X uses 0.91 kWh per gallon at 30°C and 80% relative humidity on 2026 shipping hardware, so an illustrative 42 gallon per day target is about 38 kWh per day, roughly 1.3 times what an average US home consumes in total (865 kWh per month, EIA 2024). Cooler or drier air raises it, to directionally about 2.7 kWh per gallon in Phoenix and 3.7 in Denver. Size the array to your worst month using NREL PVWatts, not to your annual average.
Can you run an atmospheric water generator on solar without a battery?
With a small one, often yes; with none at all, no. Every machine needs a modest electrical buffer for compressor startup, controls, pumps, and passing clouds. What you may not need is a large bank sized to run the machine around the clock, because a water tank can carry the household through the hours the machine is off. Whether that works at your site depends on your array, your inverter, and your climate together.
How do you make water from air without electricity?
With a passive process: a sorption hydropanel using sunlight as heat, a radiative condenser rejecting heat to the night sky, or a fog net. SOURCE, the largest hydropanel maker, publishes 2 to 5 liters per panel per day and recommends a two-panel array as standard, so roughly 4 to 10 liters daily (SOURCE published specifications, via World Water Reserve, 2026). Against WHO guidance of 5 to 7 liters per person per day for drinking and cooking, that covers one or two people, not a household. Note that no electricity does not mean no energy: every passive method depends on an energy gradient. A compressor machine, which makes household volume, cannot run without electricity at all.
What does a whole off-grid water-from-air system cost?
Four purchases, not one: the machine and its install, a solar array, a battery, and a water tank. The machine and whole-home install run $10,000 to $25,000, and the full ownership model is in our cost-per-gallon piece. Water storage runs about a dollar a gallon before installation. Array and battery sizing, which is what drives their cost, is in our solar and battery sizing piece, and it depends heavily on which of the three dispatch architectures you pick: running on the sun with a large tank moves money out of the battery and into storage, and running around the clock does the reverse.
How do I know if an atmospheric water generator will work at my property?
Run four checks. Pull your monthly climate normals from NOAA's 1991-2020 US Climate Normals for the station nearest you. Identify your worst month, usually January. Test it against the roughly 13°C and 25% relative humidity derating point for current-generation hardware. Then pull your worst-month peak sun hours from NREL PVWatts. If your worst month sits below the floor for most of the day, water from air is steeply reduced production in that month, so plan a second source or a larger buffer for it.
Should I buy a bigger battery or a bigger water tank?
For anything beyond a day or two, the tank, because a battery does not do seasonal storage at any price. Atlanta's modeled seasonal buffer of 3,929 gallons at a 42 gallon per day target represents about 3,575 kWh of electricity already spent and banked as finished water, the energy equivalent of roughly 264 Tesla Powerwall 3 units, and batteries also self-discharge and age on the calendar while stored water does not. Buy battery for the hours: compressor startup surge, controls, pumps, cloud bridging, and running after dark. Buy tank for the weeks and months. The limit is that a tank substitutes for battery capacity only, never for array capacity, because a gallon of tank helps only if the array and the air can produce the surplus gallon that fills it.
What is the minimum temperature and humidity for an atmospheric water generator?
Roughly 13°C and 25% relative humidity as a practical floor for current-generation hardware. Below either, output is low or zero and no storage medium compensates, because there is no surplus to store. The floor is a daily condition rather than an annual average. In Aquaria's capacity dataset, Miami is above the floor 24 hours a day in every month, Houston falls to 10.5 hours a day in January, Atlanta to 3.5 hours in December and zero in January, and Denver is at zero from November through February (NOAA 1991-2020 normals).
Can I just use a dehumidifier?
Not for drinking water. A dehumidifier produces condensate with no air-intake filtration, no disinfection, no food-grade wetted path, and no mineralization. The condensation physics overlap with an atmospheric water generator; the safety engineering is the entire difference.
About the Author
Brian Sheng is Co-Founder and CEO of Aquaria. He trained as an engineer at Princeton, where he studied water technology, and is a co-inventor on Aquaria's atmospheric water generation patent application (USPTO App. No. 64/009,115). He writes the technical reference layer for atmospheric water generation, including Aquaria's capacity-by-climate dataset and cost-per-gallon model.
Related Reading
- Off-Grid Water From Air: The Complete System
- How Much Water Do You Need Off-Grid?
- Can an Atmospheric Water Generator Run a Whole House?
- Solar and Battery Sizing for Off-Grid Water Production
- Atmospheric Water Generator Capacity by Climate
- Atmospheric Water Generator Cost Per Gallon
Sources
- NOAA National Centers for Environmental Information, US Climate Normals 1991-2020 (the reader's own site data, and the basis of Aquaria's capacity dataset). https://www.ncei.noaa.gov/access/us-climate-normals/
- NREL PVWatts Calculator (the reader's own peak sun hours by address). https://pvwatts.nrel.gov/
- NREL Annual Technology Baseline 2024, Residential Battery Storage (round-trip efficiency and service life; no cost figures used). https://atb.nrel.gov/electricity/2024/residential_battery_storage
- Cole, Wesley, and Akash Karmakar, “Cost Projections for Utility-Scale Battery Storage: 2023 Update,” NREL, 2023. https://www.nrel.gov/docs/fy23osti/85332.pdf
- LFP cycle life and depth of discharge: manufacturer published specifications, RELiON, 2025. https://www.relionbattery.com/knowledge/can-lithium-iron-phosphate-batteries-save-you-money-over-time
- University of Georgia Extension, Climate and Agriculture in the Southeast, “What is the diurnal cycle?” 2016. https://site.extension.uga.edu/climate/2016/10/what-is-the-diurnal-cycle/
- US Energy Information Administration, average monthly residential electricity consumption, 2024. https://www.eia.gov/tools/faqs/faq.php?id=97&t=3
- SOURCE (Zero Mass Water) published Hydropanel specifications: 2 to 5 liters per panel per day, two-panel standard residential array, 4 by 8 feet, 340 lb, via World Water Reserve, “Source Hydropanels Review,” updated March 2026. https://worldwaterreserve.com/source-hydropanels/
- Global Water Intelligence, published criticism of hydropanel economics.
- WHO Regional Office for South-East Asia, Technical Notes for Emergencies No. 9, “Minimum water quantity needed for domestic uses” (WEDC, Loughborough University).
- Aquaria capacity dataset (first-party): modeled HydroPack X output and productive hours by market and month.
Subscribe to our Newsletter
Get the latest updates, news and insights directly to your inbox
No payments for 6 months. Get your Hydropixel delivered free in the continental US.
Get Your Free Water Plan
Two quick steps. We'll send the right system size, expected output for your climate, and exact pricing — usually within one business day.
