Automatic Ice Maker - Using Pitcher Pump bug

[mathmanican]

If you think the aquatuner takes too much power, just build 3 pitcher pumps instead.  In the following example, 6000kg of crude oil dropped from 38.9C to around 20C in 1 cycle.  An aquatuner running at full power would have only dropped the temp to 24.8C. 

This bug has been around since at least Feb 2 (posted to steam, not klei forums). I found it by accident, but wrote a lengthy comment on a recent bug report (expanding the bug by a bunch).  I hope that if we exploit this bug in many different builds, the devs will take action to eliminate it. 

So, for my first exploit, here is an automatic ice maker. It can be as big, small, as you want. You could use an automatic sweeper to scoop up the ice, if wanted. Using a hydrosensor on the water layer could make sure you don't inject too much water all at once.  More tinkering could get a constant supply of ice being made.  Then conveyors can take the ice whereever you want it.

Also, we see clearly that hot things move right, and cold things move left. 

Have fun exploiting this bug. I'm excited to see the results. Currently I'm envisioning a PW boiler that uses a steam turbine to boil the water, batteries to cool the steam to 75C (exploit), then pitcher pumps to get the temp down to whatever you want. Two doors, placed below each pump prior to building the pump, would make an easy on/off switch for the cooling power. 

[KittenIsAGeek]

I've noticed counter-current heat transfer works very well in other aspects as well.  I used it in one of my early boiler systems: Cold polluted water flowing counter-current to the near-boiling pure water as it condensed.  I'll have to test again to see if "heat to the right" made it work.

[Argelle]

3 hours ago, mathmanican said:

Also, we see clearly (..)

No, I do not yet see clearly

Temperature view would help.

In the bug report you state it's not working with water, but here you transform it into ice?!

Anyway, I would not mind more explanations please.

[Mutineer]

If I understand picture correctly, there is a layer of water on top of petrol, which physically is not correct. In real-world water heavier than petrol.

[Argelle]

I'd love to understand ONI before real world. Too complicated, too much exploits in real world

[Yoma_Nosme]

Do the pitcher pumps have to be used to cool the liquid or just sit there?

[PolloFiero]

32 minutes ago, Yoma_Nosme said:

Do the pitcher pumps have to be used to cool the liquid or just sit there?

It’s a bug they found. If you place a pitcher pump on top of crude oil, the temperature will slowly go down, so it’s essential for this build.

[Oozinator]

7 hours ago, mathmanican said:

This bug has been around since at least Feb 2

[mathmanican]

8 hours ago, Argelle said:

No, I do not yet see clearly

Temperature view would help.

In the bug report you state it's not working with water, but here you transform it into ice?!

Anyway, I would not mind more explanations please.

The pitcher pump cools the oil. The oil cools the ice.  The two layers of oil are where the magic happens. Currently I'm working on a build to cool a geyser with zero power input.  The key is to to make sure the oil occupies the two layers that you see above.  Then put water on the layer above (faster than using metal tiles to transfer the heat).

7 hours ago, Yoma_Nosme said:

Do the pitcher pumps have to be used to cool the liquid or just sit there?

Yep. Without the pitcher pump, the liquid won't cool. 

10 hours ago, KittenIsAGeek said:

I've noticed counter-current heat transfer works very well in other aspects as well.  I used it in one of my early boiler systems: Cold polluted water flowing counter-current to the near-boiling pure water as it condensed.  I'll have to test again to see if "heat to the right" made it work.

The "heat to the right" part isn't what makes this work, rather just an interesting observation about heat flow. 

4 hours ago, Oozinator said:

The bug was only reported on Aug 7 in the forums here.  It's been in the game since February (based on the post in Steam). It hasn't been confirmed yet (only reported recently), but I do hope to see that change soon.

If anyone wants my save file to play around with, here it is (though I've let so many cycles pass that all the old tests are now in the -20's or -30s. Just clear out any old oil, and brush some new oil in. It's pretty easy to replicate this in survival.  Just build several pitcher pumps next to each other in oil, and you'll see the temp drop fast. Want to cool the oil biome and turn it into a new ice biome?  Just spam pitcher pumps.

Paradise2.sav

 

[Argelle]

So, in bug report it was oil and petroleum, but it works with oil and oil?

That's nice to report this, hope it gets fixed (and that we play with it in the meantime )

[mathmanican]

Here's a preview of the geyser cooler.   The input water source is >98C. The output depends on the length of the trail. I'm trying to get the right length, and avoid cyclic fluctuations in output temp.

2 minutes ago, Argelle said:

So, in bug report it was oil and petroleum, but it works with oil and oil?

That's nice to report this, hope it gets fixed (and that we play with it in the meantime )

The most cooling happens with oil, so any layer that is large enough to cover 2 tiles is best (I currently just brush in 1000kg of oil, and it takes up two tiles). It might actually work better (have to try this) to use a thin layer of crude oil (10kg), topped by a think layer of petro (10kg), then let water flow on top of that. Fun bug to play around with.

[Denisetwin]

49 minutes ago, mathmanican said:

The two layers of oil are where the magic happens

so you have one normal layer of oil with a very thin layer of oil on top of that and then a thin layer of water?

 

[Saturnus]

1 hour ago, mathmanican said:

It might actually work better (have to try this) to use a thin layer of crude oil (10kg), topped by a think layer of petro (10kg), then let water flow on top of that. Fun bug to play around with.

Tried that already. Well, with grams of each. -40C in a few seconds. Then the oil froze and stopped the process.

To have a useful output I'm experimenting with having a running petroleum flow over a crude oil layer. Seems promising.

[NanoD]

What I see in my testing is that the more crude oil you have the more cooling you will have.

Well maybe the correct is that the more the pump access the more cooling there will be.

[Saturnus]

8 minutes ago, NanoD said:

What I see in my testing is that the more crude oil you have the more cooling you will have.

Yeah. To a certain extent.

I've made a constant flow cooler that has one tile height of crude oil (870kg each), 10 pitcher pump length, able to cool petroleum flow at 10kg/s from 20C to -3.8C constantly. That's almost 24C delta with full pump flow. Quite impressive.

Also found that abyssalite pumps works best, and temp shift plates are banned as they normalize the temperature which makes the temperature dip initially but then goes back up.

 

[NanoD]

How much heat deleted in about 1 cycle. At a temperature of 30 C.

1600 kg = 4,7 MJ

3200 kg = 4,8 MJ

4800 kg = 5 MJ

[KittenIsAGeek]

I am aware that the counter-current flow is not the key.  My question, however, was if counter current flows work best if the heat is flowing to the right.

[mathmanican]

Thanks all for the further testing.  I'm going to now make a self-cooling plastic factory, hopefully one that can also produce ice from the generated steam, along with the plastic. The idea, just drop 1kg/s of petro onto a pitcher pump situated to the right of the plastic plant, and the let it flow left over the plastic mill. Maybe someone will beat me to the punch. Then I'm off to build a temperature sensor for temps up to 600C, using the infinite cooling to keep one side of the temperature gage at -20C or so. So many new uses. 

[Saturnus]

33 minutes ago, KittenIsAGeek said:

I am aware that the counter-current flow is not the key.  My question, however, was if counter current flows work best if the heat is flowing to the right.

The difference is very minor. The oil temperature gets significantly lower with co-current flow but the heat exchange is better with counter current flow. You can see a previous set up I did here. The top one has an output petroleum flow at roughly 0.2C and the bottom one is a roughly 0.1C. Almost no difference in the output but the oil is 0.1C in the top one but -2.8C in the bottom one in the coldest tile.

That was before I found out that anything else than a single tile depth of oil but a full tile (870kg) will have significantly lower efficiency as shown in the previous post where the same pump length but one tile high gets about 4C lower output temperature, or more than 20% better efficiency with 870kg instead of 1740kg.

[mathmanican]

Got the plastic mill up and running.  It works like a charm.  The crude oil is two layers, but the top layer is less than 1kg of oil (raised to full height by petro on top of it).  The petro from the left tank is pumped to the right at 1kg/s (could be less). The petro in top tank, at 98C, is pumped to the plastic mill at 833.3g/s, keeping the tank in the plastc mill empty (otherwise the middle tile doesn't stay at below freezing).  All generated steam gets turned into ice.  The long this thing runs, the more cooling power it will have. 

I didn't let it run for tons of cycles, but a temp sensor, placed under the left pitcher pump, could turn the mill and pumps off/on, if the temp of the petro in left tank ever drops too much. The let the pitcher pump get stuff cool again, and start it up all over.

Fun times.  I'll have to do the temp sensor another time. 

[Saturnus]

Btw, to those wondering if you can temperature control the pump cooler, the answer is: yes,you can! Build the row of pumps on top of doors (you need to build the doors first) that open with a temperature sensor. You should probably only do this as a fai safe option if your set up is really sensitive to temperature fluctuations as it doesn't always start back up again without issues (but it always shuts the cooling off).

[fishpear]

Glad that you popped my bug post up and it got widely discussed.

In my later experiments, the fastest cooling structure is like Saturnus's design, which is over 10x faster than my original design. 

And the air is actually not needed. Leave the upper space vacuum you can see higher cooling speed

16 hours ago, Saturnus said:

 

 

 

[mathmanican]

@fishpear I'm glad it got noticed too. This has been a fun bug to play with.

Figured I'd work on oxygen production. Who needs wheezeworts anymore. I'm guessing you could easily cool enough O2 for any number of dupes with a setup that builds off something like this. Just put the thermostat at whatever temp you want to swap which direction air flows. Pop a pressure sensor on the outlet side of the O2 to sense for the right pressure at which to turn off the oxygen production. The crude oil has a min temp it will reach, so it's not crucial to thermally isolate it from the entire system, though that could also be done with standard vacuum door locks if needed. If you don't like the liquid max pressure bug, you could easily do something similar and just replace the wheezewort setup with a pitcher pump.  Takes a tad more vertical space than wheezeworts, but puts out a lot more cooling power. 

@Saturnus I also tried something to turn the cooling power on/off with two doors below the pump, but would rather not deal with the unpredictability of what happens each time the cooling turns on/off.  Also, it takes way longer to prime, as the door mass needs to be cooled as well. 

[The House]

@mathmanican  Trying to make your build in game.  What liquid are you using with the electrolyzer?  I tried using petroleum and it keeps getting flooded.  Looks like a cool way to separate the hydrogen though.

[mathmanican]

You can use polluted water/water, or crude oil/petro, or crude oil/water, or any combination as long as polluted is not on top (otherwise it will off gas. They key is to make sure there is not enough to over pressurize the electrolyzer. You can read more about it on many posts like the one below. It's been around for a really long time.

I used crude oil and petro. To get the right amounts, I use a simple room to bottle the right amount of liquid to start with. 

Whenever I need to make a liquid lock, or get enough liquid to prevent over pressure messages, I just have the bottle emptier first drop crude oil, then petroleum.  Mopping them up gives me three tiles with just a little over 340 grams of both crude and petro. When I'm ready to build the electrolyzer, I'll first put a bottle emptier in there, make is "sweep only" and then select my crude (all three of these I think I put in that one), and then select the petro after the crude got delivered. I used to build a complicated pipe system that could deliver any specified number of grams, but the bottle emptier is way easier. 

Hope this helps.

On another note, the pitcher pump bug/exploit/feature doesn't appear to be long lasting. Any time you save/load, or disable/enable, the pitcher pump, then the cooling starts up again.  However, after a few cycles have elapsed, you may not see any more cooling power. Over the last few days, my huge oil well has gone from 110C down to 60C (in the oil biome), but most of the cooling happens in the first few minutes after each save/load.