A YouTube video from a Florida maker called Hyperspace Pirate went viral last month. The concept: use solar panels to freeze a bucket of water during the day, then melt that ice at night to cool your house. Hackaday picked it up. Interesting Engineering covered it. The comments sections exploded with people declaring this the future of air conditioning, the death of batteries, and the cheapest cooling hack ever invented.
I watched the video. I read the articles. I read every comment. And I have a perspective on this that almost nobody else in the solar industry can claim, because ice thermal storage is not some futuristic curiosity in my world. It is a technology that my own extended family helped pioneer, that I have a direct professional connection to through one of the oldest solar manufacturers in America, and that I have watched succeed and fail for different reasons over the course of my career.
Let me explain.
What Hyperspace Pirate Actually Built
The setup is straightforward. Three 100-watt solar panels charge a lead-acid battery. When the battery is full, a microcontroller triggers a small compressor that freezes a two-gallon bucket of water. The compressor uses R600 (n-butane) as a refrigerant. Once the water is a solid block of ice, insulation keeps it frozen for several days. When cooling is needed, a separate glycol loop circulates through copper tubing embedded in the ice block, picks up cold, and pushes it through an automotive radiator with a small fan. The fan and pump draw only a few watts.
He tested it in his truck. The system cooled the cab noticeably in about two hours. The discharge rate was roughly 700 watts of cooling, comparable to a small window unit, delivered with almost zero electricity at the point of use.
As a proof of concept, it works. Nobody should argue otherwise. The physics are real. Water has a latent heat of fusion of 334 kilojoules per kilogram. One cubic meter of ice stores roughly 93 kilowatt-hours of cooling capacity. That is a real number with real applications. It is also the exact reason commercial ice storage has been a legitimate technology for decades.
But there is a canyon between a proof of concept that cools a truck cab and a system that reliably and economically air conditions a Florida house. That canyon is where every residential ice storage idea has gone to die.
This Is Not New. Not Even Close.
Ice storage air conditioning has been around since at least the 1970s in commercial applications. The concept is simple: run chillers at night when electricity is cheap, freeze water in a tank, then use that stored cooling during the day when rates are high. The economics work because commercial buildings often face demand charges and time-of-use rate structures that make peak daytime electricity extraordinarily expensive.
CALMAC, a company founded in New Jersey, introduced its IceBank energy storage tank in 1979. It became the gold standard for commercial ice storage almost immediately. The system uses glycol circulating through polyethylene heat exchangers inside insulated tanks. Freeze at night, melt during the day. Over 4,000 buildings in 60 countries use CALMAC thermal storage, including Rockefeller Center, the Credit Suisse Building in New York City, and Goldman Sachs (which famously has 92 ice tanks under its trading floors, holding enough ice for 3.4 million margaritas, according to Bloomberg). Trane acquired CALMAC in 2017, recognizing that thermal energy storage was not going away. In fact, the global ice thermal energy storage market was valued at $3.1 billion in 2024 and is projected to reach $8.3 billion by 2033, growing at over 11% annually.
That is not a failing technology. That is a thriving commercial technology. The distinction matters, because every time someone tries to shrink it down to residential scale, things fall apart.
My Father-in-Law Holds Two Patents on This
Ed O’Neal is a mechanical engineer who has served Southwest Florida for over 50 years. He is one of the most accomplished HVAC engineers in this region. He holds two United States patents related to thermal energy storage for air conditioning: US Patent 5,372,011 and US Patent 5,860,287.
The first patent covers a system that uses a fixed-capacity compressor to simultaneously charge both a cold storage tank and a heat storage tank using phase change materials. The goal was elegant: instead of cycling a compressor on and off (which wastes energy, shortens compressor life, and causes temperature swings), you run the compressor continuously to build up stored cooling and stored heating, then use small pumps to deliver that stored energy to the space as needed. The compressor shuts off for extended periods while the stored energy does the work. Think of it as a thermal battery for your air conditioner.
The second patent refined the coolness storage side specifically, focusing on how to efficiently freeze and discharge a phase change material to deliver sustained cooling from a fixed-capacity compressor without the cycling problem.
Both patents were technically sound. The problem was not the physics or the engineering. The problem was the economics and the timing. To make the system work at residential scale, you needed insulated storage tanks, phase change materials, dual heat exchangers, circulation pumps, solenoid valves, thermal expansion valves, bypass circuits, pressure sensors, and a control system to orchestrate everything. You were starting with a cheap compressor and surrounding it with a small chemical plant. The installed cost would have blown past the price of simply buying a better compressor.
And that is exactly what happened in the industry. Inverter-driven variable-speed compressors became affordable and widespread through the 1990s and 2000s. They solved the cycling problem far more elegantly. The compressor ramps up and down to match the load. No tanks. No phase change materials. No extra pumps. One component replaced an entire subsystem.
Ed specified CALMAC tanks on commercial jobs throughout his career. He saw firsthand that the technology worked beautifully at scale, where the economics of peak demand charges and time-of-use rates justified the capital investment. At residential scale, a clever solution got overtaken by a simpler one. That is not a failure of imagination. That is just how technology works sometimes.
Ice Storage Is Already Here
Southwest Florida is no stranger to ice storage systems.
If you drive east on Ben Hill Griffin Parkway toward Florida Gulf Coast University, the cooling for 26 campus buildings comes from ice. Not metaphorically. FGCU operates one of the largest ice thermal storage plants at any university in the country. Over 200 CALMAC IceBank tanks sit behind the central energy plant, freezing water overnight using FPL’s off-peak rates and delivering chilled water through two miles of underground pipe during the day. The university shuts its four chillers off entirely between noon and 9 p.m. every day during the cooling season. FGCU saves over $400,000 per year and has among the lowest energy costs per square foot of any university in Florida. The system has been in place since the campus opened in 1997.
Drive south to Shell Point Village, the largest continuing care retirement community in Florida, and you will find another CALMAC installation. Shell Point’s ice storage system won a 1996 ASHRAE Technology Award for innovative and efficient cooling technologies. CALMAC lists Shell Point alongside Goldman Sachs, Rockefeller Center, and Google on its featured installations page. When the University of Arizona was planning what became one of the largest ice storage systems in the world, they sent a team to Fort Myers to visit the Shell Point installation as a reference site.
My father-in-law, Edward O’Neal, P.E., is a founder and principal at Wadsworth O’Neal Associates, an MEP consulting engineering firm in Fort Myers that has been designing mechanical systems for commercial, institutional, and medical buildings across Southwest Florida for decades. He did not read about ice storage in a magazine or watch a YouTube video. He designed it, patented variations of it, and put it into buildings that people in this community walk through every day without knowing what is underneath them.
The FAFCO Connection You Did Not See Coming
Here is where the story gets personal in a way I did not expect when I first saw the Hackaday article.
Before I started Florida Solar Design Group, I worked for a FAFCO dealer right here in Southwest Florida. FAFCO, Inc was the oldest and largest solar thermal manufacturer in the United States, founded in the 1960s. They built their reputation on polymer heat exchangers for solar pool heating, panels that sit on your roof and circulate pool water through small, closely spaced tubes to absorb heat from the sun. Those panels are on thousands of roofs across Florida. I sold them. I installed them. I know the product inside and out.
What most people in the solar pool heating world do not know is that FAFCO also manufacturee ice storage equipment. FAFCO SA, the European arm of the company based in France, produced the IceBat Cold Thermal Energy Storage system. The IceBat uses a patented glycol-filled polypropylene heat exchanger to freeze water in storage tanks and deliver cooling on demand. The same polymer heat exchanger technology that FAFCO developed for solar pool panels found its way into commercial ice storage. Over 1,400 IceBat systems are operational worldwide, delivering more than 3.6 gigawatt-hours of installed cold thermal storage capacity. A hospital in Paris saved 53,000 euros per year by freezing ice at night (using negative nighttime electricity tariffs) and supplying chilled water during the day.
So the company that taught me solar thermal technology was also one of the most successful ice thermal storage manufacturers on the planet. The polypropylene panels I used to carry up ladders and bolt to roofs share direct DNA with the heat exchangers inside commercial ice storage tanks cooling hospitals and office buildings across Europe. Solar thermal. Ice storage. Glycol circulation. These are not separate worlds. They are the same world, built on the same heat transfer principles, by some of the same people.
When I watched a Florida YouTuber circulate glycol through a homemade ice block and push cold air through a radiator, I was not watching something revolutionary. I was watching a simplified version of technology that my father-in-law patented, that the company I used to work for sold, and that has been cooling commercial buildings since before I started my career.
Sadly, FAFCO closed its doors in 2024, ending a storied history in the solar thermal world.
Why Solar-Powered Ice Storage Does Not Work for Homes
The Hackaday comment section had it half right. Several commenters pointed out the obvious: batteries exist, and they power more than just your air conditioner. But the deeper problems with residential solar ice storage are worth spelling out, because they are the same problems that have killed every residential ice storage attempt for the last 30 years.
The efficiency penalty is real, and it is significant. Making ice requires cooling water below 32 degrees Fahrenheit. Your air conditioner normally cools air to around 55 degrees at the evaporator coil. Pushing a compressor to freeze water instead of just cooling air means a larger temperature differential between the evaporator and the condenser. Larger temperature differential means higher refrigerant pressure differential. Higher pressure differential means the compressor works harder and draws more power. Carnot efficiency sets the theoretical ceiling, and real-world systems fall well below that ceiling. One commenter put it simply: increase your ambient temperature 30 degrees and your efficiency drops roughly in half. Making ice in the Florida sun is thermodynamically worse than just running your air conditioner directly.
The scale problem is brutal. A typical Florida home needs 3 to 5 tons of cooling capacity. One ton of air conditioning is historically defined as the heat needed to melt one ton of ice in 24 hours, which is 12,000 BTU per hour. To carry a 3-ton home through eight hours of no-solar cooling (evening through morning), you need roughly 288,000 BTU of stored cooling. That is about one ton of ice. Not a two-gallon bucket. One literal ton. 2,000 pounds. About 240 gallons of water. That is an insulated tank the size of a large hot tub sitting somewhere on your property, plus a dedicated chiller to freeze it, plus a glycol circulation system to discharge it, plus controls to manage the charge and discharge cycles.
The cost comparison with batteries has shifted dramatically. When Ice Energy launched the Ice Bear for residential applications in 2014, lithium-ion battery costs were still high, and ice storage seemed like a cheaper way to shift cooling loads. But battery prices have fallen by roughly 90% over the last decade. A Tesla Powerwall 3 holds 13.5 kWh of usable energy, costs a fraction of what it would have a decade ago, and powers everything in your house, not just air conditioning. You can run your lights, your refrigerator, your internet, your medical equipment, and your air conditioner off the same battery. An ice tank only stores cooling. It does nothing for any other load in your home. And when you add up the cost of the storage tank, the dedicated chiller, the glycol system, the controls, the insulation, and the installation labor, you are not saving money over batteries. You are spending comparable money for a single-purpose system.
Ice Energy learned this the hard way. The company, which was perhaps the most visible residential ice storage advocate in the country, filed for Chapter 7 bankruptcy in December 2019. They did not fail because the technology did not work. They failed because the economics could not compete with rapidly falling battery prices. A reinvigorated version of the company is now focused on commercial applications in California, which is telling. The residential dream is gone.
The Carnot Problem Nobody Wants to Talk About
Here is the part that gets overlooked in every enthusiastic ice storage discussion. When you use solar panels to run an air conditioner directly, the energy conversion chain is short: sunlight hits the panel, electrons flow, the compressor runs, your house gets cool. Two conversions. Solar to electric. Electric to cooling.
When you use solar panels to make ice and then use that ice to cool your house later, the chain gets longer: sunlight hits the panel, electrons flow to a battery (charge/discharge losses), the battery powers an inverter (conversion losses), the inverter runs a compressor (which must work harder because it is freezing water, not just cooling air), the compressor freezes water (more losses because you are pushing to a lower temperature than needed for direct cooling), the ice sits in a tank (thermal leakage losses through insulation, no matter how good), then a pump circulates glycol through the ice (pump energy), and a fan pushes air across a radiator (fan energy). Every step in that chain loses energy.
A lithium-ion battery has a round-trip efficiency of 85% to 95%. An ice storage system, when you account for the thermodynamic penalty of freezing water instead of directly cooling air, the thermal leakage during storage, and the pump and fan energy to discharge, has an effective round-trip efficiency well below that. You are burning more solar energy to deliver the same amount of cooling.
If you have unlimited free solar energy and no other use for it, the efficiency losses do not matter. But that is almost never the case. Every kilowatt-hour of solar energy has value, either as electricity you use directly, electricity you store in a battery for later, or electricity you export to the grid. Turning it into ice is the lowest-value use of that energy in almost every scenario.
Where Ice Storage Actually Wins
None of this means ice storage is a bad technology. It means it is a technology that works in specific contexts, and residential solar is not one of them.
Commercial buildings with demand charges and time-of-use rates can save real money by shifting cooling production to off-peak hours. The CALMAC systems that my father-in-law specified on jobs are still running, still saving money, and still being installed in new buildings. The payback periods for commercial ice storage are often three to five years, which is excellent for a capital investment.
District cooling systems in hot climates, particularly in the Middle East and Southeast Asia, use massive ice storage to serve entire neighborhoods from centralized plants. The scale justifies the infrastructure cost.
Grid-scale thermal storage is growing as a way to absorb excess renewable generation. Wind farms produce the most power at night when demand is lowest. Freezing ice with that excess wind energy and using it for daytime cooling is economically rational at utility scale. This is one of the fastest-growing applications for ice thermal storage worldwide.
Facilities with critical cooling needs, like hospitals and data centers, use ice storage for resilience. If the chiller goes down, the ice tank provides hours of backup cooling without any electricity. That is a value proposition that batteries can match, but at a higher cost per unit of cooling.
In every one of these cases, the common thread is scale. Ice storage gets cheaper per unit of stored energy as the system gets bigger. Water is almost free. The tanks are relatively simple. The glycol loops are standard HVAC components. What kills it at residential scale is that all the ancillary equipment (controls, pumps, valves, insulation, dedicated chiller) represents a fixed cost that does not shrink proportionally with the tank size.
The Florida Residential Reality
In Southwest Florida, where I design and install solar energy systems, the cooling season is essentially year-round. We do not have the luxury of making ice in a cool winter and using it in a hot summer. The outside temperature is working against you when you need ice the most. A compressor freezing water when the ambient temperature is 95 degrees with 80% humidity is fighting thermodynamics the entire time.
FPL does not offer time-of-use rates for residential customers by default, so there is no rate arbitrage incentive to shift cooling to off-peak hours. The primary economic driver that makes commercial ice storage work does not exist in the residential context here. You are making ice with expensive daytime solar electricity (which has an opportunity cost because you could have used it or exported it) and then using it at night when grid electricity in Florida is not particularly expensive.
Meanwhile, a solar panel system with battery storage lets you capture daytime solar energy, store it, and use it to power everything in your home at night, including the air conditioner. The battery does not care whether it is running a compressor, a refrigerator, or a light switch. It is infinitely more flexible than a tank of ice.
If you are building an off-grid cabin on one of the barrier islands we serve by boat, where every watt matters and every piece of equipment has to be barged in, the calculus does not change. Battery storage still wins because it serves every load, not just cooling. The simplicity advantage of batteries over a dedicated ice storage system is overwhelming in a remote installation where maintenance access is limited.
What the Commenters Got Right
The Hackaday comments were surprisingly thoughtful by internet standards. Several people correctly identified that ice storage is a proven commercial technology being reinvented by someone who may not know the history. One commenter who works in Florida commercial HVAC pointed out that ice storage makes economic sense when you can make ice during off-peak hours and discharge during peak hours, which is exactly how it has been done in commercial buildings for decades.
Another commenter nailed the battery comparison: lithium-ion batteries lose 5% to 15% in a charge-discharge cycle, while ice storage through the full chain of freezing and discharging loses considerably more. And the battery powers everything, not just cooling.
What the commenters mostly missed is the personal connection that runs through this technology in Southwest Florida. My father-in-law patented thermal storage systems for air conditioning. He specified CALMAC tanks on commercial projects. The company I worked for before starting Florida Solar Design Group was a dealer for FAFCO, which manufactures both solar pool heating panels and commercial ice storage systems. The polymer heat exchange technology in a FAFCO solar pool panel and the polymer heat exchange technology in a FAFCO IceBat ice storage system are branches of the same engineering tree.
I am not watching this from the outside. I grew up professionally in the intersection of solar thermal energy and thermal storage. That intersection is real, it is commercially successful, and it is completely wrong for residential solar applications in Florida.
The Bottom Line
Solar-powered ice storage for residential air conditioning is a technically valid concept that fails the economic and practical tests for homeowners. The physics work. The engineering works. But the cost, complexity, efficiency losses, and single-purpose nature of an ice storage system make it a poor investment compared to battery storage, which has plummeted in price and serves every load in your home.
Commercial ice storage is a different story entirely. It is a mature, growing, multi-billion-dollar industry that reduces peak demand, lowers operating costs, and increasingly helps integrate renewable energy into the grid. CALMAC (now Trane), FAFCO, and others have proven the technology in thousands of installations worldwide.
If you are a homeowner in Southwest Florida thinking about how to store solar energy for use at night, the answer is a battery. If you are a commercial building owner or facility manager thinking about how to reduce demand charges and improve cooling resilience, ice thermal storage deserves a hard look.
And if you are a YouTuber in Florida freezing a bucket of water with solar panels, I respect the ingenuity. I have been surrounded by people doing exactly this kind of work my entire career. Just know that two patents, a bankrupt company, and 50 years of mechanical engineering history suggest the residential version is not going to replace what is already sitting in your garage, and storage solutions exist today that cost less and do more.
