Berkeley Lab scientists develop a new method of refrigeration | Hacker News Reader

Berkeley Lab scientists develop a new method of refrigeration

newscenter.lbl.gov

139 pointsgmays40 comments

From the article: >The first experiment showed a temperature change of 25 degrees Celsius using less than one volt, a greater temperature lift than demonstrated by other caloric technologies.

This makes no sense. Voltage is not a measure of energy, it's a measure of electrical potential, or electromotive force. Low voltage with extremely high current is still a lot of energy. The only thing important with refrigeration efficiency is how much (electrical) energy it requires to achieve a certain change in thermal energy; for this, you'd need to measure the voltage and current, or just the wattage, plus measure the actual thermal energy change (usually measured in BTU for air conditioners).

The article is a bit confusing on that point. From looking at the paper it seems like the 1 volt is relevant not because it's supposed to be implying energy efficiency by itself, but for some other reasons I don't understand.

From the abstract:

> "Our experimental results show a coefficient of performance of 30% relative to Carnot and a temperature lift as high as 25°C using a voltage strength of ~0.22 volts."

And from the actual paper:

> "To modulate the electrochemical potential in a real system, the ion concentration can be controlled by applying a voltage in an electrolytic cell (e.g., dual-ion battery), where the applied voltage is typically ~1 V. This stimulus is considerably milder than those used in magnetic, electric, and pressure-based caloric systems."

I'm not following the paper at all, but it sounds like requiring a low electrical potential is a good thing.

Electrolysis of water becomes extremely favorable above 1.23 volts, and is not something you want to subject most electrodes to. The closer you get to 1.23 volts the more and more likely it is for electrolysis to happen at a given current used to maintain the field they'd need.

I'm not 100% on this given I don't want to buy the paper and am not an expert in this field, but have done a lot electrolysis related work on personal projects.

edit: Also electroplating would probably become a concern as well depending on what exactly is in that brew.

> Voltage is not a measure of energy, it's a measure of electrical potential, or electromotive force

I don’t understand the difference between electromotive force and energy associated with electricity. How can there be low electromotive force with high energy?

I'm not an electrician, but drawing an analogy to water you can think of voltage as water pressure in the sense that you can have a small stream of high pressure water as in a water pik or a large current of low pressure water as a river. The river carries more energy despite having a lower voltage.

Exactly: this is the normal analogy used for understanding voltage and current. In a pipe with water, you could have a gargantuan pipe carrying water from a lake or river, at very low pressure (only gravity is moving it), but with a gigantic volume, so the total energy is huge. Or you could have a tiny pipe in your house with water at high pressure (60-90 psi is normal for municipal water supplies), but the total amount of water is low, so the total energy is low compared to that huge pipe with river water.

The analogy breaks down once you reach sonic flow in the pipe however.

It can be bs. However it's still interesting because a lot of refrigeration uses a compressor which requires a lot of voltage (and current) to start up. It might be like peltier coolers which are solid state, but unfortunately not very efficient and don't cool much in practical terms. (search thermoelectric coolers or peltier)

One interesting cooling technology is the free piston stirling cooler. Coleman used to sell one years ago for ~$400, and then discontinued them. They are highly prized now. Basically they took very little energy to start cooling something and could cool to very low temperatures. The coleman model seemed to use CO2 as the cooling medium. They still sell similar coolers now, but for >$2k usually with helium as the medium. They are so pricey now they are mostly used for medical transport. There might be some "you need a license" stuff with these types of coolers since they can go quite far below zero

I've also noticed you can buy the FPSC piston part itself (without the box) for random prices. search "free piston stirling cooler" on amazon.

It's too bad they didn't take off, I think they would pair very will with solar for solar-powered refrigeration.

>However it's still interesting because a lot of refrigeration uses a compressor which requires a lot of voltage (and current) to start up. It might be like peltier coolers which are solid state, but unfortunately not very efficient and don't cool much in practical terms. (search thermoelectric coolers or peltier)

This is exactly why it's BS: a compressor requires high(er) voltage to start and run, and Peltier coolers can run on 5V USB outlets. However, compressor-based cooling is FAR more efficient than Peltiers, which are infamously inefficient. Voltage really doesn't tell you anything. The problem is that journalists have no idea what a Volt is.

They are not comparing to compression cycle efficiencies. They are comparing ionocaloric results to other caloric cooling techniques.

Electrocaloric cooling requires an E field of hundreds or thousands of volts per cm to cool a material 15K or so, so it's remarkable that ionocaloric cooling can be driven by a weak field. They never imply that it is more power efficient than every other cooling technique -- that idea was injected upthread.

It does take energy to create and sustain such a strong field. But in any case, fields that strong break polymers and other materials (air) and are harder to work with generally. EC/MC cooling devices are expensive and inefficient as a result.

> journalists have no idea what a Volt is

Institutional science communication doesn't work like that. Bylines aside, these articles are an agglomeration of what the researchers, communication team, grant managers, lab directors, and DOE bureaucrats want to say.

I don't think I understand either, however, with thermocouples it sort of makes sense - increasing current also increases ohmic heat, so per volt performance is important.

It is strange but perhaps this is referring to a ratio. A temperature change similarly does not imply any energy delta (in classical physics at least) as the mass of an object approaches zero.

I agree though, just tell us how much more efficient it is than existing technology.

>> Low voltage with extremely high current is still a lot of energy

To be a bit pedantic, this is strictly a lot of power - it is necessary to multiply by time to get energy.

1 amp * 1 volt = 1 watt = 1 joule / second

But you can draw conclusions based on that. Low voltage also means that the amount of electricity is limited - the lower the voltage, the thicker a wire you need for any serious amount of power. In practice, it means that 1 volt will be using power equivalent or below e.g. regular USB chargers.

I mean sure, voltage was not the right term to use, but it doesn't necessarily mean they cheated.

From the paper: "In electrocaloric devices, electric fields of ≈200 MV m^−1 are common (more than 50 times as large as the dielectric strength of air) and can only be generated by applying voltages across micrometer-sized films in the kilovolt range. ... In the ionocaloric device, the applied field is ~0.22 V, which is relatively modest. ... For additional perspective, some high-efficiency aqueous vanadium redox flow batteries operate at 800 mA cm^−2; the ionocaloric device operated at ≈0.5 mA cm^−2 for its highest-efficiency results. If, for example, it operated at 800 mA cm^−2, like in the aqueous system, the cooling power output would be equal to 9.2 kW L^−1."

It's fairly common for articles like the one linked to be written by the lab's communications department, who may or may not be familiar with the subject at hand.

It's one of those journalistic, press-release kind of write-ups, the kind that spend some effort explaining basic concepts in a hand wavy kind of way, before proceding to pitch the main point with yet more hand waving interleaved with quotes from those involved. I think the point may be just to make the public at large feel positively about the New Thing.

You are partially right regarding energy, but not all joules are created equal. Being able to use a small differential such as 1 V is an achievement in itself.

I haven't studied the new method in detail. But I will say they have an annoyingly poor understanding of the current state of vapor compression system refrigerants.

Because from the technological standpoint this is a solved problem. Domestic refrigeration runs on hydrocarbons (your fridge is likely running on pentane). Supermarket refrigeration runs on CO2. Both of these have proven efficiency, safety and environmental aspects, with millions of units installed. Large commercial refrigeration (slaughterhouse etc.) runs on hydrocarbons or ammonia, the latter is of course toxic, but in a large factory you have adequate procedures to handle that safely.

The two biggest challenges wrt. refrigerants today are to ensure that the old and bad solutions are phased out worldwide and not just in the richer countries, and to make sure the same transition happens for heat pumps, where the R1234-variants are being pushed now but are still problematic.

Just created an account to post exactly this. How such a statement could make it into the article is really baffling.

Of the most efficient heatpumps in Europe, most are running on propane, which has negligible GWP ( global warming potential https://hydrocarbons21.com/ipcc-includes-gwps-for-hydrocarbo... ) see for example https://lambda-wp.at/ (only in German, sorry)

Also, air conditioners for the private home can run on propane , see https://www.green-cooling-initiative.org/news-media/news/new...

And here they've demonstrated an approach which uses a material that is zero GWP, non-toxic, non-hazardous, etc. I get that fridge techs are happy with the way things are, but research continues regardless.

One thing I have to note and keep wondering about is whether these "Berkeley Lab scientist", "Google researcher", "Institution X" add anything of value to the actual content of the research? A publications uses to the name to add credibility, the institution itself wants its name there even if it is not associated to the research (as was the case for the "google researcher solves math problem" one), but do we really need that on HN?

It does sound like a weird appeal to authority, now you mention it. Here's the actual paper with names: https://www.science.org/doi/10.1126/science.ade1696

Berkeley Lab employs both authors of this paper. Berkeley Lab, and this research, gets its funding from the US Department of Energy. One of the authors is a pretty high-level manager, of the sort who drives strategic planning there. It's safe to say the institution is associated with this research.

I do personally regard national lab affiliation to be a meaningful quality signal -- at least strong enough that if I'm interested in the topic I'll read the actual paper, which isn't something I can say for a lot of stuff coming out of certain universities lately.

Another good reason to note the institution is that this is a demonstration that DOE Office of Science research funding is generating results.

The DOE Office of Science funding is generating papers as results. If my interpretation is correct, the technology discussed is _potentially_ useful.

To me the downside of marketing results aggressively to justify one’s funding is that the quality signal can become inversely correlated. Is this valuable research or is someone who knows to play the DOE game inflating a metric to secure their next round of funding and/or expand their turf? Does this negatively impact researchers who do not have access to a comparable marketing apparatus?

They clearly describe having built an experimental device and outline expectations for improvement of the process. All technology is "potentially useful," until someone uses it.

As far as "marketing," the labs are contractually required by the government to do this. Average 'impact rating' and so forth are part of the performance evaluations. As far as "playing the DOE game," there are a lot of voices in the critical path of getting significant funding from the Office of Science, many of them generally healthily skeptical. I'm not aware of very many charlatans achieving high-level management positions or controlling significant funding.

Your employer basically owns everything you do. Working at a big tech company, you can't contribute to open source without permission of your employer.

In the "Google researcher solves math problem" article, the researcher would have had to get permission to publish the work. If Google had wanted, they could have kept the result secret and used it internally.

No, that's actually incorrect. They can request that sure, and the guy might comply, but it's against the law and easily contestable. These clauses are invalid. But it's even more absurd to bring this up as a reason why WE should comply with it.

Without access to the paper, and only (derivatives of) obfuscating university press releases, the supplementary material [1] allows fumbling together a naive idea of what this is about.

On page 13 a model system is provided with a "SI.6 Thermodynamic Analysis of Ideal Ionocaloric Cycle with Regeneration" diagram to stare at. Alternative substances to ethylene carbonate are described on page 5. Sodium iodide is selected as the salt that is needed in high concentration "to move the phase boundaries".

The mechanism of regeneration seems to be electrodialysis [2], which "is used to transport salt ions from one solution through ion-exchange membranes to another solution under the influence of an applied electric potential difference", a word completely missing in TFA.

There's a nice exploded view of the "Figure SI.10 Electrodialysis Experimental Setup" on page 23, after which the specific Nafion NRE-212 cation exchange membrane [3] and Fujifilm AEM Type 1 anion exchange membrane [4] are listed.

The 2015 article "Present and future caloric refrigeration and heat-pump technologies" [5] is listed as reference, which gives an overview of different "caloric energy conversion" methods and states that "elastocaloric refrigeration represents the most promising alternative, and magnetocaloric refrigeration is a very promising alternative for future applications."

[1] https://www.science.org/doi/suppl/10.1126/science.ade1696/su...

[2] https://en.wikipedia.org/wiki/Electrodialysis

[3] https://www.h2planet.eu/nl/detail/NafionNRE212

[4] https://www.fujifilm.com/nl/en/business/manufacturing-proces...

[5] https://www.sciencedirect.com/science/article/pii/S014070071...

If I recall correctly, reducing greenhouse gas emissions from refrigerants is one of the top opportunities highlighted in Drawdown [1].

[1] https://drawdown.org/the-book

Are leaky refrigerants still as big of an issue? CO2 is fairly benign, ammonia... is toxic but doesn't it break down over time? Not sure on that one.

Anyway, I'm sure the combined electricity usage is a bigger issue than the refrigerants. If this (and other) technologies can help reduce that, that would be great.

Mind you, there's a few things that can already be done to reduce power usage of cooling solutions; better insulated buildings (with shutters, awnings, and heat-reflecting windows), chest fridges, evaporative coolers, etc.

There was a video about this clay fridge (https://mitticool.com/product/mitticool-clay-refrigerator50-...) that cools its contents (by 10-15 degrees compared to room temperature) just from evaporating water; analog tech like that could be used anywhere to help cool things down.

fwiw, most people know "clay fridge" by the name "zeer pot". pretty easy to DIY.

>This could give us a place to use CO2 from carbon capture.

You don't need ten trillion tons of ethylene carbonate for refrigerators, unfortunately. Or, um, fortunately.

Isn't this "ioncaloric" process the process that makes a propane flame refrigerator work?

Oh, they say:

> No one has successfully developed an alternative solution that makes stuff cold, works efficiently, is safe, and doesn’t hurt the environment,”

OK, so a propane fridge emits GHG when it burns the propane, and ammonia is not the safest compound in the world. Got it.

This work is interesting, but the claim unreasonable.

So now we have the acoustic cavity, this, and that weird optical thing I read about over a decade back (I dimly recollect that it was light-activated and there was a strange glass involved).

I always get a kick out of refrigeration, and I think I know why: you're straight up fighting entropy (locally of course) in one of its most raw and obvious forms.

Sometimes I wonder how many basic methods there are and if they can be combined in bizarre, almost unfathomable ways.

There is also: https://en.wikipedia.org/wiki/Magnetic_refrigeration which I periodically check in on to see if their is any progress or commercialization, but so far it doesn't seem like it.

> I always get a kick out of refrigeration

It's a cooler subject.

(sorry)

has already been discussed here: https://news.ycombinator.com/item?id=34156227

Wow don't see an article like this every day? At a wedding over the summer met someone who is a refrigeration researcher and wasn't previously aware of this topic of study, nice to gain slightly more context into the field

Have seen half a dozen articles about this, all of them sketchy press pieces that leave me wondering.

It sure would be nice to see some diagrams.

The article makes it sound like it's little more than theory at this point.

The key phrase is right in the first sentence: "could someday [...]" Usually (and especially in this case) a strong indication that reading the article is a complete waste of time.

Reply on news.ycombinator.com