Just add water: Efficient Energy’s HFC-free chillers arrive in the UK An unusual data centre cooling system that dispenses with expensive, environment-damaging refrigerants in favour of tap water is now available in the UK. The eChiller, developed by German startup Efficient Energy, is built on a somewhat ridiculous premise: instead of fluorinated gases like R-134a – which have a high global …
@Luke Maslany
We have a Datacentre modelling application, that allows us to place hardware from a catalogue into racks, and work out hot spots and air flow etc. Most kit is in the catalogue, but sometimes I have to go look for a spec sheet, and then the heat output can either be BTU or KW. The App defaults to BTU, because we're British, and awkward.
The only place I've ever heard of anyone specifying Aircon units in BTU or tons in actual seriousness in the last 40 years has been the USA - Portugal is a new one (presumably they're just using american glossies)
Most places use kW(latent) and kW(sensible) - the difference being critically important when you find some twat's installed comfort cooling in a server room as only the (sensible) part matters.
Latent cooling is about removing moisture from the air, reducing humidity and allowing us bags of "mostly water" to sweat, thereby feel cooler. Reducing humidity does little for a heatsink's ability to transfer heat unless you've had a sudden conversion to swamp coolers.
It's worse than that, because many comfort cooling systems will just freeze up in a server room. I recall explaining this to someone who had exactly this problem, and were at first a little sceptical that an aircon unit would freeze up because the air is not wet enough !
Some systems are specified to work under any humidity conditions - these are OK as long as you size the unit accordingly. However, many systems rely on the heat taken into the evaporator coil from the latent heat of all that condensing moisture to keep the coil above freezing point. So in humid conditions they work fine. Put them in dry conditions, the latent heat is missing, so the coil gets colder - whatever moisture there is in the air then freezes on the coil and unless the unit has frost detection and automatic defrost, the coil gets blocked with ice. Of course, when the coil is (partially or wholly) blocked with ice, the unit will still be trying to remove heat, so that ice gets very cold so it takes some shifting.
A ton is simply 12,000 BTU, so it's a pretty straightforward conversion. Generally tons are used when it's bigger than a domestic system, which would be measured in BTU. A house might have 24,000 BTU while a commercial chiller could have five tons, rather than say 60,000 BTU.
But what's missing from the article is EER/SEER, which are the ratios of BTUs to watts. How efficient is the system? If it needs more electricity, then its GWP is worse. Refrigerants may not be good for the environment but these are sealed systems, and it normally gets collected when the system is removed or replaced. The worst refrigerants are no longer in use anyway, except (illegally) in China.
"The App defaults to BTU, because we're British, and awkward."
Speaking as a 44 year old British engineer with a mixed heritage of structural and thermal design, I only use kW. In my experience, it's *mostly* the US that uses BTUs, but there seems to have been a resurgence in making calculations more complicated than they need to be over here recently...
I thought cooling was measured using BTU
Technically it's measured in BTU/hour, which is dimensionally the same as Watts. 1 BTU is 1055J (or 252 calories if you prefer), so 1 kW is 3412 BTU/hour.
The reg standards convertor seems to lack units for energy and power, so the best real unit I can come up with is that 1 BTU/hour is .02 NorrisLinguini/second
(Mines the one with a very old "Data and Formulae for Engineering Students" in the pocket.)
For an odd historically-based value of 'logical'..
"Logical" as in "logic says you wouldn't want your primary ports shelled by dreadnoughts until they were smoking rubble, would you, so do go along with our units like a good chap".
I've never understood why the Yanks used Imperial measures after the 1790s though, given that they sided with the French Republic.
Yes, I went to live in the UK before supermarkets went metric.
112(!) pounds to a hundredweight, 14 pounds to a stone (which is only used to weigh persons), 16 ounces to a pound, no obvious link between liquid and fluid ounces.
A Gill which is a unit you divide to sell booze in pubs (different fractions depending on geography); a furlong which is some distance a horse is running; feet were used for vertical distance but yards for horizontal.
Different sizes of gallons depending on geography.
It was probably good for my mentaller arithmetic when I went shopping, though.
14 pounds to a stone (which is only used to weigh persons),
And potatoes!
That's really why there was such resistance to the change. The metric system is clearly better for calculations, scale drawings, etc. but in other cases it wasn't the units that mattered, but what they meant. My Grandmother would have known exactly how many people she could feed with "half a stone" of potatoes, but "3kg" would have been meaningless to her in that context. Likewise for distance, tell someone that something is "15 miles" away and they don't think of the linear distance, they automatically think "a short drive" or "too far to walk". When you're used to that mental image, "24km" doesn't bring the same instant comprehension. It takes time to readjust.
no obvious link between liquid and fluid ounces.
1 fl.oz of water weighs 1 oz
Different sizes of gallons depending on geography.
Because of the difference in underlying pint sizes. USAnians get shortchanged on their beer.
That reminds me of one of the objectors to metrification in Denmark (1906 I think): "The system should only be introduced when all who had trained under the old system have retired."
Regarding ounces: Yes, almost - density is =.9978 ~ depending on temperature. May have mixed up with troy ounces.
That's really why there was such resistance to the change. The metric system is clearly better for calculations, scale drawings, etc. but in other cases it wasn't the units that mattered, but what they meant.
I was having a discussion about metric/imperial things with a mixed group of European engineers having seen something in France which was clearly labelled in inches. A German in the group said his father was a carpenter and that wood is measured there in imperial cross sections but metric lengths, so he'd use 2M lengths of 4x2 to build things.
I also noticed recently when replacing plumbing parts in my posh German shower that it used 3/4" fittings.
"14 pounds to a stone (which is only used to weigh persons),"
A butcher's stone was 8 pounds.
Just when you thought you'd gotten your head around it all.
As for the distances or weights - I lived through the metric conversion in au/nz. ('73-75)
As kids we simply became bilingual in measurements (except fahrenheit - those are still alien).
Anyone ten years younger speaks metric-only, anyone ten years older is bilingual in decimal currency/£sd too.
The UK is metric in all but name. The road signs - posted in miles - are at km and metre multiplier intervals (and the engineering ones ARE in metric). The USA is the same.
The issue isn't that that metric is "different", but that under the old system there are _too many_ different units of measurements with strange multipliers between them. Rods, chains, perches, poles, yards, paces, feet, inches, miles, furlongs, leagues, fathoms, knots being just one set of examples.
And then there's the problem that there are different measurement units WITH THE SAME NAME - 6 different competing pounds/ounces over recent history. 2 actively used different measures with differing pints, fluid ounces, quarts and gallons.
And to cap it all off - going further back EVERY COUNTRY had its own unique sets of measurements. You can't standardise for widespread commerce in that kind of environment and whilst a country _might_ be large enough to try and say "our measures are good enough for us", this becomes at least as large a barrier for selling things OUTSIDE the country as for anyone on the outside attempting to sell things into the country.
You can see a similar effect with cars - there are three world car standards - UN(LHD) UN(RHD) and USA(LHD) - USA carmakers are now crying it's "unfair" that they are required to conform to UN(LHD) standards to sell their vehicles in the rest of the world and that the rest of the world should conform to USA(LHD) standards - which were mostly enacted as way of erecting a trading barrier to imports without breaching GATT rules in force at the time and had the effect of creating a captive market.
Most recently the US government has been trying to make out that requiring vehicles comply with "rest of world" safety standards before being sold in 3rd party countries is an illegal tariff barrier and they will take punitive action against countries that block imports on that basis. I wouldn't be at all surprised if they start pulling this against countries that drive on the left hand side of the road and require that the steering wheel be on the right hand side of the car (for obvious safety reasons)
And of course it wasn't so long ago that if you serviced cars you needed at least 4 different sets of tools to handle the differing standards used by the manufacturers (Acme, Whitworth, UNC, UNF, 4 different UK types, metric, etc etc etc)
"16 ounces to a pound"
Only for the Avoirdupois pound(0.45359237kg - or 7000 grains)
It's 12 for a Troy pound or Apothecaries pound (0.3732417216kg - 5760 grains)
It was 12 for the Tower (Hill - site of the royal mint) pound (0.350kg - 5400 grains)
and 12 for the London Mercantile pound (0.437 kg - 6750 grains)
The wool pound was 6992 grains - and wasn't divided
The Pound of Sterling (a Tower pound) was equal to 240 sterling silver coins - which means a silver penny had an approximate purchasing power around to 60p today. No wonder there were fractional values - all the way down to a groat.
Why 12? Romans! This is all descended from the roman Libre - which was divided into 12.
Why grains? (and they're the exact same grain used to measure gunpowder and other items) - They're grains of barley, which are remarkably consistent.
Why 16? Who knows? It's French! (So was the Troy pound)
The really fun part is that Apothecary pound is divided into 12 ounces, which are divided into 20 pennyweights or 8 drams. Drams are further divided into 3 scruples - so when someone has "no scruples" or "only had a dram"....
Why are fluid ounces different?
So 1 pound H2O 1 deg F warmer?
Strangely simple and consistent compared to other imperial luminaries like stones (14 is such an under-appreciated prime-based multiple: of course being merely 1 stone overweight sounds waaay better than 14 pounds fatter), miles (5280 feet - what’s the likelihood Joe Schmoe remembers that?). 2pdr guns (do they change the caliber for DU rounds?). Fahrenheit (0 F is coldest ever, uh wait and 96F is body temp, darn 98.6), alcohol proofs (gunpowder playing a prominent role in drinks, somehow). Gallons - 10 pounds of H2O at 62 F.
Knots, because why count sea speed the same as land speed?
Still, air velocity is being left out, lacking the recognition of its own unit, so I propose a new unit, the Swallow. Care should be taken not to name the species so that Imperial Swallow != Merkin Swallow, unlike the fiasco whereby UK mile == US mile (thank Deity for Short Tons vs Long Tons and different UK/US gallons).
IBM has used water cooling for many systems in the mainframe and supercomputing space for years, but it required a sink for the heat.
In the 9125 F2C supe. there was captive water circulating around the frames, cooling chips and power converter components either directly or via heat pipes, together with mid-plane air heat exchange units for the remaining components. The 'water' had several additives for anti-corrosion and gas quenching (to reduce cavitation), so you would not want to drink it, although it was non-poisonous and not corrosive (but still had to have hazardous substance labeling because of H&S regulations).
But they needed to be plumbed into a cold-water supply that could take 100s of kilowatts of heat away from the frames through water-to-water heat transfer devices (called Water Conditioning Units in IBM speak), and in the case of the systems I looked after, this went to the refrigerant units for the air conditioning, although I was assured that any suitable water source could be used
These systems were so efficient that according to the machine room manager, the air that was pushed through the frames for the non-directly cooled components apparently came out colder than when it went in. It gave them quite a headache balancing the cooling for the rest of the room, as it was not directly controlled by the AC, and the amount of cooling from these systems actually depended on the compute load on the cluster!
That's apparently why Einstein and Szilard invented a fridge. Ammonia used to be used in domestic fridges and fatal accidents happened.
> There are already refrigerants with low or zero global warming potential that are suitable for industrial use (eg ammonia).
Adsorbsion systems have their own sets of problems. SolarFrost have nailed down a lot of them over the last 20 years but many remain (including overall efficiency at power levels below 500kW - not a problem if you're using waste heat or solar power but otherwise it can be a serious issue)
For that kind of setup I'd keep the ammonia side of things outside and run a chilled water/antifreeze loop inside the data centre. Ammonia, biologicals and and enclosed spaces are not a good combination.
Agreed - as the required final temperatures is well above freezing using chilled water inside the data center is the best approach. (Even better if the racks have built in cold water channels), Large quantities of ammonia (or most other refrigerants) are best kept outside where a leak is less of a problem. A leak on a chilled water line is less of a problem and cheaper to fix than a leak of any refrigerant gas.
the UK is expected to leave the EU, possibly as soon as October 31, and those nasty refrigerant gasses could be about to stage a comeback.
or not.
'New UK ODS and F gas regulations transfer most of the requirements of the EU regulations into UK law.'
that brexiters have been loudly proclaiming their right to scrap EU rules
Correct.
including environmental protections
Citation, please.
In practice the most likely victims are rules affecting farming and land use.
The ones requiring farmers to set-aside land to keep prices high, you mean?
By and large the UK tends to have better (and admittedly often excessively gold-plated) consumer protection law than UE minima. Nothing so far suggests that Brexit will change that.
Surprise, surprise but that goes for many other EU countries too.
Why should that be a surprise? It simply demonstrates that in order to keep 28 countries on-board the EU has to aim for the lowest common denominator. Most countries indeed can do better than its rules & red tape when left to their own devices.
Or not. Neither of the two giants, China and the USA have a remotely good record in food or industrial regulation, particularly when it comes to safety.
And they are certainly excellent examples of self-governing proudly independent of God or Man countries left to their own devices; and one day they will kill us all.
For thous Brits who want to know more about food safety in the USA and the pros and cons of mega food production there is the American made story, Food.inc
https://www.youtube.com/watch?v=C4osE1BjdPw&t=633s
And should the Brits after a brexit go to the US for a deal, on their knees feeling they stand tall and equal there is of course the pesticide Roundup to enjoy.
https://www.youtube.com/watch?v=2l-EzOyxh6w
As far as I know only the French has so far said "no" to that stuff in the EU.
pesticide Roundup...As far as I know only the French has so far said "no" to that stuff in the EU.
No, they keep kicking that can down the road in the face of considerable protests by farmers. It's still allowed.
Then again, it's France, so even if it were banned people would keep on using it.
"they keep kicking that can down the road". Compared to who, the British?.
What the Wikipedia has on this is:
"Ban in France
In January 2019, Roundup 360 was banned in France following a Lyon court ruling that regulator ANSES had not given due weight to safety concerns when they approved the product in March 2017. The ban went into effect immediately. The court's decision cited research by the IARC, based in Lyon.[52][8]".
the UK is expected to leave the EU, possibly as soon as October 31, and those nasty refrigerant gasses could be about to stage a comeback.
If the UK really were as idiotic as our current so-called Government seems to be, it quite probably could repeal that legislation, but it would find that international manufacturers would still be making cooling systems to be compliant with the requirements of a large 510 445 million population sales market, rather than bother to cobble together any separate substandard models for the minority population of a fast-decreasingly-significant post-imperial (in all senses) country determined to push itself into its own decline.
And, if the UK still has any manufacturing industry left by the time it may or may not throw itself out of the EU, our manufacturers, if they have any sense, would also continue to only manufacture products to comply with the requirements of the larger market. Even if they could then legally get away with shabbier products for Grot Britain, it still wouldn't be worth the effort of having separate design and manufacturing lines.
it still wouldn't be worth the effort of having separate design and manufacturing lines.
Precisely, which is why the "sky might be falling after Brexit" attitude is just more tiresome, glass-half-empty, FUD. FFS guys, don't you have any ambition to do better than the minimum?
"if the UK still has any manufacturing industry left by the time it may or may not throw itself out of the EU"
At this point, with the increasingly violent rhetoric and talk of "traitors" being tossed about I wouldn't be at all surprised if many of the larger companies contemplating moving out are planning ways of organising shutdowns for "extended maintenance", then moving in a fleet of trucks, packing up the critical kit and only announcing they're going AFTER those vehicles have safely arrived on the other side of the Chunnel.
I'm also surprised that more recruitment posters for jobs in mainland europe haven't started popping up. I'm certainly seeing more headhunting going on.
The UK was in a rotten state before it joined the EU, losing population at a prodigious rate and only bailed out via being a member plus a lot of oil/gas revenue (mostly pissed against the nearest wall by various governments). The oil/gas ran out some time back and the UK's been a net importer for a while. The lack of forward planning since 1980 shows and trying to blame Johnny Foreigner for the actions of sucessive incompetent governments has gone down as well with Johnny Foreigner as might well be expected (ie: they've rolled their eyes, watched the poo-flinging tantrum, told us we still won't get any extra ice cream and we're perfectly welcome to leave home to go live with the bears in the woods, but they don't eat porridge and we will go well with ketchup.)
If the UK really were as idiotic as our current so-called Government seems to be, it quite probably could repeal that legislation
Except our government isn't, attempting, to do... anything.. of the sort.
PARLIAMENT at the will of voters in the future may do although it seems remarkably unlikely given precisely zero people are asking for it.
I see we've still conveniently forgotten that the UK is a world leader in this stuff, no thanks to the EU.
We need to ban H2O! For the children, of course. Or, at the very least, require a warning label for it. If you breath it, you will die. If you drink too much of it, you will die. At just the right temperature, it will scald your skin. Long-term exposure to the skin has been proven to cause temporary skin deformities. We need the nanny-state to protect us from this dangerous H2O!
This post has been deleted by its author
I can find nothing about the workings of this specific super efficient product online. Not even a block schematic on the web site.
Does this mean it's just a couple of heat exchanger systems with one mother of a cooling water supply, or perhaps a huge radiator with a fan? Or perhaps it's a block of peltiers stapled to a water cooler block and a radiator loop which would probably work well but be as electrically inefficient as hell ... I'd like more info ...
I think it must run at very low pressure, so that the boiling point of the water is close to or below ambient - then you can use fairly conventional expansion valves and compressors to move the heat around.
But yes, it would have been nice to get a bit more substance in the article.
It does indeed rely on evaporation/condensation cycles. Some explanations and a sketch of the chiller there:
"It does indeed rely on evaporation/condensation cycles. Some explanations and a sketch of the chiller there:"
A closed-cycle version of a cooling tower with a vacuum to help things along on the evaporation side. It looks typically German engineering (i.e. more complicated than it needs to be) since it's not getting much of a thermal advantage over the performance of a cooling tower.
This post has been deleted by its author
It needs both.
Compressor to return the low pressure steam to water, pumps to move the liquids around.
(Nerd history note - the turbine on the Titanic and some other ships of the era worked entirely below atmospheric pressure. The outlet from the reciprocating engines was somewhat below atmospheric, and the turbine exhausted into vacuum. It's surprising how much energy there is in large volumes of low pressure steam.
Working below atmospheric made the turbines safe using the metallurgy and seals of the day, so it was very much a win/win situation.
AIUI, the best refrigerants rely on phase transitions to help transfer the most heat. In terms of water, it deals with the most heat in the Heat of Vaporization/Condensation (2257 J/g absorbed to evaporate, emitted to condense, assuming sea level). It takes about four times the Joules to boil off a mass of water than it does to take it from freezing to boiling points. And one wonders why steam is so dangerous.
So the system has to be configured so that the water is in a pressure state such that by the time it's absorbed enough heat, it's evaporated into steam. Then, it has to be condensed to such an extent that by the time it lets off the heat, it condensed back into liquid water.
The older refrigerants were easier to manage these transitions at common conditions.
I used to work at a facility, a wind tunnel, that had to dissipate MW of heat. This was done using evaporative cooling - similar to the cooling towers seen at power stations but in a more compact form. The same systems are available as HVAC systems that can cool from an ambient of 35C to 25C. These systems work well but take a lot of maintenance. The water has to be purified, filtered and treated to prevent legionella.
Here's some manufacturer's bumf on the subject.
Government owned-and-operated local test facility had a cooling pond as a cooling fluid (mostly primary; sometimes secondary) for the dynamometers, power converters, and even the engines/powerpacks/vehicles under test. Pond surface was about 1 to 2 acres, if not larger** -- more like a lake! -- and the temperature was near-constant as long as it wasn't frozen. They filtered and treated it to not corrode or clog anything up.
** I last saw it full-size about 10 years ago. First they reduced it to build an addition to the main test building. Now it looks like it's gone entirely in lieu of a test track. It's difficult to remember the original borders.
What it looks like today (and where it is): https://goo.gl/maps/1zTRa5NUYMLRYj7G6
Eleven/11 Mile Road* actually stretches quite far across Metro Detroit, much of it as a service drive for Interstate 696; look again closer in along the freeway. The name comes from it being 11 miles north of the grid's center in downtown Detroit: https://en.wikipedia.org/wiki/Mile_Road_System_(Michigan)#Metropolitan_Detroit
* There is zero consistency in using the number or the word for any of the mile roads, unless it's a half-mile. I work off 18 1/2 Mile and live between Hall Road, which is technically 20 mile, and 21 Mile.
This post has been deleted by its author
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
