Isaac Solar Ice Maker | Hacker News Reader

Isaac Solar Ice Maker

energy-concepts.com

83 pointsabrax314146 comments

My non-electric refrigerator operates along similar lines but uses a propane burner to dissociate ammonia from water in a continuous cycle. Devices such as this can operate on many different heat sources including kerosine. Absorption refrigerators have been available for over a century but they are neither cheap nor inherently safe. https://en.wikipedia.org/wiki/Absorption_refrigerator#Single...

My next off-grid refrigerator will be solar-electric with a high efficiency compressor. Even factoring in the cost of additional solar panels and a storage battery it will be cheaper to purchase and certainly cheaper to run.

What’s not safe about them? I thought that absorption refrigerators were developed to be safer than existing refrigerators, before non-toxic working fluids were discovered.

Einstein-Szilárd fridge. Yes, that Einstein and that Szilárd. https://en.m.wikipedia.org/wiki/Einstein_refrigerator

Ammonia is a very dangerous chemical. From a thermodynamic and environmental perspective it's one of the best refrigerants for many applications but we can't use it in residential buildings because you'll die if you breath it in.

The E-S refrigerator was used because it's high reliability made it safe for certain critical applications, not because it used a safe working gas.

> Ammonia is a very dangerous chemical. From a thermodynamic and environmental perspective it's one of the best refrigerants for many applications but we can't use it in residential buildings because you'll die if you breath it in.

FWIW I’ve only seen propane/ammonia cycle fridges inside residences so I don’t understand what you mean by “can’t use it in residential buildings”

Scale. A small personal fridge in an RV or small apartment isn't a big deal. But a system big enough to power comercial refrigeration or air conditioning to a residential building will involve enough coolant that any leak would be serious problem.

Yeah, but indirect (with e.g. water/glycol between the condenser and the indoor air handler) systems can be combined with scrubber tanks (basically just water+baffles, I think) that the blow-off from a leak or a pressure relief valve is forced through, in order to contain most of the ammonia.

Many supermarkets use ammonia refrigeration. Many food warehouses and packing plants use it as well- resulting in incidents fairly often. It's still used.

Few people sleep in supermarkets.

Anhydrous ammonia is the favored refrigerant for big commercial and industrial systems. These need professional management anyway, and the plumbing containing ammonia tends to be outdoors. Gaseous ammonia is buoyant, so goes up.

The amount of anhydrous ammonia that would be needed to operate an RV fridge could kill you easily. Operation in an enclosed space is another source of risk.

Excellent point, and the difference here is really interesting! Residential refrigerators use a mixture of 90% ammonia and 10% water. This is not very dangerous at all, but it also isn't nearly as energy efficient as pure ammonia because it has different thermodynamic properties.

Pure ammonia is dangerous because it has an extreme affinity for water and it will dissolve into any water it comes into with. If you breath in ammonia gas, it will dissolve in the mucus of your lungs and draw in more water from the surrounding tissue, flooding your lungs and killing you. If the ammonia is already mixed with a little bit of water, this isn't an issue (hence why you can have ammonia cleaning products in your home).

All that said, I was wrong about E-S chillers. I did some more reading and both E-S and absorption chillers use hydrous ammonia.

The burner is potentially problematic. Inefficient combustion can lead to carbon monoxide, a poorly maintained burner can start a fire. Gas leaks, kerosene spills, etc. Electric refrigeration is much safer.

https://www.cpsc.gov/Recalls/1998/cpsc-warns-that-old-servel...

That's a recall from 1998. Surely the appliances produced now are both safer and more efficient?

At my offgrid cabin I converted a $300 chest freezer to a fridge by using an external thermostat that toggles power off/on based on temp. I have it turn on at 37 degrees and turn off at 35. It uses such little power that it's not even a factor for my small solar setup.

Chest freezers are great because they dont have a front-facing door which lets all the cold fall right out.

Indeed. Not as easy to organize as one with a front facing door, but we keep stuff organized in milk crates and it's worked great for our family for many years now.

The New Yorker recently published an article on the scarcity of refrigeration in the developing world:

"Africa’s Cold Rush and the Promise of Refrigeration

For the developing world, refrigeration is growth. In Rwanda, it could spark an economic transformation."

https://www.newyorker.com/magazine/2022/08/22/africas-cold-r...

I'm reminded of what happened to solar hot water. It's now generally cheaper to use PV to drive a heat pump water heater. I suspect the same is true here -- use PV to drive an ordinary refrigeration cycle. And PV, unlike the concentrating collectors shown here, also works with sunlight scattered off clouds and clear air.

Cheaper to install or cheaper to run? My understanding was that the solar heaters were relatively cheap up-front.

Cheaper to install and also to maintain.

Depends on type of solar heaters -- vacuum insulated ones used in colder climates are not that cheap

Sure, but unlike PV "The construction of the ISAAC Solar Icemaker involves only welding, piping and sheet metal work, and there are no expensive materials."

So this can be produced and maintained/repaired with much simpler tools, that makes it much more feasible to use in remote area's or developing countries.

Or more so that mechanical refrigeration (as opposed to river ice extraction and shipping) may be a small scale technology that could exist past supply chain collapse

No expensive materials but it costs $7,000

Mostly labor, no doubt.

$7000/12 is $583, so not cheaper than a regular refrigerator, but it uses no fuel or electricity, which for a dozen refrigerators would run over $4000/yr.

And it uses concentrating collectors. That involves moving parts there, to track the sun. Are they assuming the tracking system is manufactured locally, or were they assuming manual tracking? Talk about labor intensive.

Alternately, observe that PV is going to be like window glass -- a product for which transport costs become very important, so it will be made in facilities close to where it is used. The 3rd world should have PV factories.

Why is it cheaper?

Rapid R&D, weird economies of scale, Chinese subsidies, dumping, etc. PV modules are very cheap these days (often < $1/watt for the modules themselves, then maybe another $1 to $1.50 for the rest of the costs). There are also installers everywhere.

For solar water heating, the nice ones use special vacuum-insulated collectors that are expensive in and of themselves, then you have to pipe all that liquid somewhere, through the roof, often to a special water heater made to accommodate a second heat source. Or have a roof strong enough to have a rooftop water tank. And you have to worry about freezing. In general these aren't common in the USA anymore, but are elsewhere (a 2012 report analyzing reasons https://www.nrel.gov/docs/fy12osti/54793.pdf)

As the PV economies of scale ramp up, it's easier to just drop that onto any house (no major retrofits needed, just a few roof holes and a quick electrician approval) and use an electric, possibly tankless, water heater to do the job. The side benefit of that is that you don't have to worry about shunting waste heat. If you generate too much electricity, you just sell it to the grid and/or put it in batteries.

You can have both, but with limited roof space, it's often easier just to get PV. If the costs & paybacks are similar, PV is a lot less maintenance than SWH and is more widely available using commodity components.

Thanks for the explanation. I suppose you could also have a DC water heater and get some efficiency from not turning the solar electricity into AC?

Yes, you could, like this: https://www.esmagazine.com/articles/102597-low-cost-direct-s...

I would say the best setup depends on your goals, e.g. "save as much money as I can" vs "be totally off-grid and self-sustaining" vs "have a superbly engineered system that I personally designed but nobody else can understand or maintain".

Starting from a finite roof area, you can install X modules. The higher-efficiency ones (meaning more power per square area) are more expensive. Then you have system efficiency losses from everything from DC/AC inversion (as you pointed out) to wiring to temperature to cloud cover to dust to sunlight angles, whatever. Ballpark, you can expect to get roughly 80% efficiency (meaning you can expect 80% of the array's nameplate output -- the sum of what your modules' wattage is listed as -- to be usable power that your loads can use). If you're going to throw batteries in there, you would lose a bit more efficiency, especially as time goes on.

Example 1: All you care about is hot water

If for some reason you don't care about the rest of your power draws in your house, you can fill your roof entirely with PV modules (generates electricity) or solar thermal collectors (sunlight heats a coolant, coolant gets pumped inside and a heat exchanger heats your water tank), then have the entirety of your production go towards your water heater. Excess heat beyond your temperature setpoint has to get shunted somewhere (so you don't boil yourself alive) and is generally wasted, unless you can heat exchange it with some other home heating system (radiant floor heating, for example). Otherwise you might just have one really hot utility room or dump the heat back outside. This is a super simple system, but you're essentially losing out on all the excess production.

Example 2: If you want to optimize payback by selling back to the grid

The most common residential PV systems are simple grid-tied ones where you generate power and buy/sell from/to the grid as needed. These days 300W+ modules are very common, so many households can generate more power on their roofs than they need, especially during the summer. Depending on your local electric utility and power buyback rate regulations and how they interact with time-of-use billing, it may make sense to either sell back to the grid all the time, or sometimes bank that excess production in a battery bank/electric car/supplemental heat for a water heater, etc. until you can sell back at a higher rate. It becomes a game of power arbitrage and you're basically a small power plant for the utility to buy from.

Example 3: You don't want to be connected to the grid at all

The math for this changes a bit because you're no longer concerned about time-of-use or selling back electricity, you probably just want the most days of autonomy you can get (i.e. how many cloudy days you can withstand before the lights go dark). That's a trickier formula involving your roof space, sum of your household average loads, your peak draws (do you need to run power tools or an electric stove/heater/air pump), the sunshine-days (average insolation) in your area, etc. From there you'd work backwards, figure out an appropriate battery bank size (due to chemistry, li-ion and lead-acid batteries have optimal depths of discharge you want to calculate for), maybe get the highest-efficiency modules you can find, etc.

Other examples: AC-coupled hybrid on/off grid systems (to survive power outages but also be able to sell to the grid), DC-direct households (like you can use DC appliances and lighting, such as those made for RVs, throughout your offgrid cabin), ground mounts to add additional PV modules if you have a lot of electric cars or whatever, tilt-mount axes (not as popular anymore because it's cheaper to just add modules), microinverters and power optimizers to maximize the per-module output on a curve... it can get super complicated if you really want it to.

But I'd caution against unnecessary overengineering.

If you go with a bog-standard grid-tied setup, not only will it be the cheapest (probably), you can more easily repair/replace components as they wear out in the future. Batteries have the most limited lifespan (10 years? maybe more if you use modern chemistry and manage it really well), followed by inverters (15-25 years?)... the modules themselves are probably going to outlast the rest of the system.

If you make something super complicated, it's going to be harder to find people who can maintain or replace it as it ages. Solar is not a especially high-margin business, so resellers and installers go in and out of business all the time, and if you have something really nonstandard, you'll have to pay extra for specialist engineers who can understand your particular setup vs any standard solar install crew.

But it's a cool field... have fun and sorry for being so longwinded, lol

In warmer countries - I've seen them in Greece and Turkey - solar boilers are on pretty much every roof, they're all collectors and a tank of I presume oil or some other medium, like on this stock photo: https://www.123rf.com/photo_16565679_solar-collector-tank-co...

I believe PV also stops working effectively at higher temperatures or more intense solar radiation; I haven't seen much solar power down there.

Yep, also, say, Israel, Tunisia, in practice most of south mediterranean countries, basically anywhere houses have terrace roofs.

The tank is just a water boiler, to store the heated water and usually (not always) it has a resistance heater for those few cases/days when hot water from sun heating is not enough.

The only issues with these is that they are suitable for single homes (or however small, self-standing,) as the water pipes have to go up to the roof and back, the commmon ones are pumpless[1] (so multi-stores/multi-apartments becomes a problem), though they do make much more complex systems (with pumps, mixers. etc.) for such buildings.

Example (from a Greek manufacturer):

https://www.andrianos.gr/en/products/solar-thermal/solar-wat...

PV's efficiency declines a bit at high temperature, or at least that of Si does. Higher bandgap materials do better (CdTe is the commercialized one).

That you have seen solar water heaters in some countries doesn't mean they're still a good idea there. The technologies have been changing rapidly.

The plus side of solar water heater is that the materials are much cheaper. In countries where there's no frost in winter it's a nobrainer.

If you find this interesting, then read the book or watch the film "The Mosquito Coast" by Paul Theroux. The whole story is about a fridge such as this operating in the heat of Africa.

> The whole story is about a fridge such as this operating in the heat of Africa.

In Honduras, Central America: https://en.wikipedia.org/wiki/The_Mosquito_Coast_(novel)

I thought that was run on hydrogen, which would be a nightmare for many reasons. Pure solar (either PV or direct heating) seems like it’s way simpler, if slightly less efficient.

Chemically/physically, how does this work? What are the inputs and outputs?

(edit: n/m, it's not generating ammonia, it's just using it as a coolant)

The patent has more detail: https://patents.google.com/patent/US4744224A/en

Input is sun and ammonia, I guess from fertilizer. Or, wait, the ammonia is part of the system and not a consumable.

Ah, thank you! I mistakenly thought ammonia was produced from the process somehow. It's not, it's just used in a closed loop as a refrigerant.

What happens when the ammonia leaks in to the river?

Its pH changes temporarily and that kills many species in the ecosystem if sustained. But it's far less toxic than other compounds or other types of spills such as oil

And they’re already dumping way more ammonia than this could leak directly into the ecosystem via fertilization. It’d take a lot of these leaking to have much impact. (And maybe the plants around it will do better for a bit when it does leak! :-)

Ammonia is also naturally occurring, from decomposition of urine.

Ammonia at standard temperature and pressure is a gas. It is poisonous to breathe if enclosed at sufficient density, but once it gets outside it's no risk to anything.

A buoyant gas. It goes up.

A potful of liquid ammonia boiling on the ground nearby can be bad, for a while.

Reply on news.ycombinator.com