How to cool water (or other stuff) with Energy as only input (the-drip-down-effect)

[Ketmol]

An experiment how to cool down water in a closed environment with energy as the only input. I use an aqua-tuner in the experiment with a drip-down-effect. The aquatuner in it self will not have a net cooling effect.

The water drip effect in a sealed environment 

 

 

 

[Saturnus]

Try building this instead. It has much higher efficiency.

The important parts as illustrated are: 1) the outer top half tiles must be abyssalite but doesn't need to be insulated, and 2) there must be metal tiles in the middle, either regular metal tiles or a sideways mechanical airlock.

Liquid can be anything but remember to set the temperature sensor to the freezing point of the liquid +20C at the start and then carefully bring it down lower if needed otherwise you risk freezing the liquid really fast because this cools down the liquid at about 2C per second.

[Ketmol]

@Saturnus I guess the metal in the middle is just for having tiles with high thermal conductivity? 

Shortly after making the video I realized that the tiles did the trick (maybe the machinery to? But defiantly the tiles. Higher conductivity of the tiles means a more effective machine. But It should in theory work with everything but abyssalite in the middle. Just not be very effective. Right?

My experiments usually take a bit of time since I don't use de-bug mode and create all in a normal game.

I did and updated version with the same principle that was in my video with in and outflow of water. Not using metal tiles, but pretty much my own design from the video (the video was recorded yesterday). It cools down about 1500g water from 60 degrees to 20 degrees per second. So that is  a 40 degree difference of 1500 g of water/second

I will try modify it with metal tiles to see if that improves it.

Would be interesting to know how much your micro machinery can process in terms of water/deegree cooling/second 

[Saturnus]

10 minutes ago, Ketmol said:

Would be interesting to know how much your micro machinery can process in terms of water/deegree cooling/second 

I can definitely test that but I might need to extend it a bit because the potential throughput is so large it can't be achieved with a single liquid pump.

Before the upgrade that gave crude oil a freezing point I used it to liquefy 2kg/s PO2 at +30C to -200C LOx using only 185W on average. So the pump/aquatuner would only run less than 1/8th of the time. 

[clickrush]

So when I understood this correctly the two builds do exactly the same, just that @Saturnus' is tighter and uses metal instead of normal tiles. This would mean that the drip down effect is dependend on the conductivity of the tile the liquid falls on?

One assumption to explain this behaviour could be that the water and the aquatuner both try to change the temperature of the tile(s) and then somehow the cooling effect wins. Since the conductivty of the tile plays a role, this could be based on neglecting a timing factor in the calculation.

edit:

This could mean that we can possibly exploit the same bug to heat things up?

[Ketmol]

Conductivity place a role here (thx @Saturnus for showing your build. With exact same setup only changing to metal plates I got mine up from lowering about 1550g/s by 40 degrees to around 2200 g/s

[Ketmol]

I am quite sure *** is the metal plates in the middle where most of the magic happens.

The major reason my machine was underperforming was that the water below did not touch the metal-plates. Once I filled up some more to touch them from below Effect more than trippled.,.... I m currently running 4kg through/s at a 40 degree temp change and the thermo aquatuner still turn of a lot from automation shutdown from to cold temp.

 

This means it can probably be further improved by using temp plates and wire bridges to transfer heat between metalplates and water.

[Saturnus]

As promised I ran a test set up with indirect cooling.

It's not a very efficient set up and can easily be improved by making it directly cooled instead but it was an interesting experiment.

The system cools 10kg/s H2O from 59.8C to 0.8C using 529KJ/cycle on average. Deducting the 2 liquid feed pumps running 100% of the time that means it's 418W(avg) for the aquatuner and pump involved in the cooling. Or the same as running 2/7th of the time.

Basically that means with further optimization it should be possible to cool at least 45-50kg/s H2O 40C delta constantly with just a single aquatuner.

[Ketmol]

@Saturnus thank you. Am I correct to assume that all this is an effect of the metal plates getting cooled down "more than they should" ( a lot more) thus removing heat. The higher the conductivity, the faster it cools down.

[Saturnus]

4 minutes ago, Ketmol said:

@Saturnus thank you. Am I correct to assume that all this is an effect of the metal plates getting cooled down "more than they should" ( a lot more) thus removing heat. The higher the conductivity, the faster it cools down.

It's high conductivity (I use tungsten above) combined with the generally known heat deletion bug where a small amount of liquid counts the same as a full tile. So the few grams on top of the metal tiles cools all liquid tiles below it is in contact with as if it was a full tile of liquid on top. Since the grams on top is constantly being replaced that means you get an enormous amount of cooling.

[BT_20]

@Saturnus with this exploit could it be possible to scale it in a way where a steam geyser can run all of the time and always be cooled to say 1 degrees c and if so how much power Would such a system use.

[Saturnus]

20 minutes ago, BT_20 said:

@Saturnus with this exploit could it be possible to scale it in a way where a steam geyser can run all of the time and always be cooled to say 1 degrees c and if so how much power Would such a system use.

Yes. Using the above horrible set up it would use 285W on average but I'm sure with a bit of tweaking I could get it well under 100W on average.

[Saturnus]

@BT_20 Here's a setup that uses 113.5W average to cool a geyser output to 20C average. I'm pretty sure it can be improved to go under 100W.

[NanoD]

A not drip cooling system. But using same bug that drip cooling use.

Not sure how many Watts it uses. But 40 kJ did cool 65 000 kg of steam geyser water to about 2-4 C. 

[FIXBUGFIXBUGFIX]

If you want to challenge to cool a geyser with as few watts as possible, one wheeze without electricity is enough.
In the beginning, everything is hot(>90℃) except the wheeze. After three cycles, the water in main tank is colder than 40 degrees.
I build them in debug mod, but it's also easy to build in a normal game. However, be careful that the wheeze must be sealed by abyssalite (normal tile is enough) to get the best effect.

wheeze water cooling.sav

[Saturnus]

53 minutes ago, R9MX4 said:

If you want to challenge to cool a geyser with as few watts as possible, one wheeze without electricity is enough.

Well. The premise of the OP was energy only. So no wheezeworts allowed. Otherwise it'd be easy to do with a single wheezewort.

[midjones]

I don't understand how you guys achieve low power consumption, given that you're using a pump plus 1000 watt aquatuner. Aquatuner only running 10% of the time via automation? Does the liquid not circulate when automation turns off your aquatuner cooling?

[Saturnus]

20 minutes ago, midjones said:

I don't understand how you guys achieve low power consumption, given that you're using a pump plus 1000 watt aquatuner. Aquatuner only running 10% of the time via automation? Does the liquid not circulate when automation turns off your aquatuner cooling?

1200W actually. Pump and aquatuner is 1440W so 113.5W on average as my example above means that measured over 5 cycles the aquatuner and pump runs less than 8% of the time, or about 47 seconds every cycle.

No. The aquatuner blocks flow when powered off.

[BT_20]

Quick question if you were to use something like supercooled hydrogen with the vertical cooling bug could you make a hydrogen bubbler with insanely broken power to output ratios.

[Saturnus]

5 minutes ago, BT_20 said:

Quick question if you were to use something like supercooled hydrogen with the vertical cooling bug could you make a hydrogen bubbler with insanely broken power to output ratios.

I've tried but haven't gotten it to work. It seems to only work well for liquids.

[donutman07]

You would have to come up with a reliable way to get a small amount of super cold gas in one tile above a large density  hydrogen tile. But unless it was like 1000kg or more it'd still be less effective than just doing it with liquids. The plus side is it's easier to have high density gas, but I'm not sure how you'd get it to touch the low density cold gas without the high density tile just spreading out. 

 

Maybe use CO2? Acts kind of like a liquid. Drip a small amount down onto a huge pool of CO2 held in place by a thin layer of hydrogen or some other gas.

[BT_20]

 

2 minutes ago, Saturnus said:

I've tried but haven't gotten it to work. It seems to only work well for liquids.

So if you were to use liquid oxygen or hydrogen would the system be able to maintain it as a liquid and keep it from boiling or freezing.Or should I look towards the door vacuum exploit to achieve such a system.

[Saturnus]

1 minute ago, BT_20 said:

So if you were to use liquid oxygen or hydrogen would the system be able to maintain it as a liquid and keep it from boiling or freezing.Or should I look towards the door vacuum exploit to achieve such a system.

The supercooler principle works for any liquid that has minimum 20K difference between freezing and evaporation points.

So yes, it works with liquid oxygen too. But not with liquid CO2 for instance.

[Ketmol]

17 hours ago, BT_20 said:

Quick question if you were to use something like supercooled hydrogen with the vertical cooling bug could you make a hydrogen bubbler with insanely broken power to output ratios.

 

17 hours ago, Saturnus said:

I've tried but haven't gotten it to work. It seems to only work well for liquids.

In theory. it should be possible to first use a hydrogen bubbler to first get liquid oxygen, then cool the oxygen using the above discussed principle with an aqua tuner even further.. then use the liquid oxygen to cool the hydrogen I guess...

I forgot to check at what temp oxygen freeze to solid thought.. but if it is lower than the point where hydrogen becomes liquid it should work....

 

[Kabrute]

The difficulty in amassing what 5 tons of liquid oxygen just to get that system started and thats assuming you precool everything first o.o I like a worthy challenge