Jump to content

The Water Duplicator - A Homemade Water Geyser Providing Almost 20kg/s at 3C


Recommended Posts

Right so I'm finished tweaking my "mini". The goals was to have 8kg/s average output at minimal power and material cost.

I discovered that doors are better for melting the ice, not because it's faster than metal tiles, as it's not, but because they consistently gives the doubling effect whereas metal tiles do not. This might be related to the fact that doors are meant to displace liquids and gases in the game in normal operation while tiles are not. They only displace liquids and gases when built. Curiously, only having doors in the middle 4 tiles seems to be a lot better than having all the bottom tiles doors.

To make this as easy and cheap to build as possible every kind of exotic material is banned from this build. Everything metal is iron or iron ore but could be copper and copper ore, or gold and gold amalgam just as well. Everything not metal is igneous rock but sandstone will work too. Granite and obsidian, not so much though. There's no temp shift plates so no diamond is used either. Cooling liquid is polluted water so apart from the fact that you need to rush a mechatronics engineer this can be built extremely fast if you have a high Learning dupe preferably with interests in research and operate that you run through as apprentice scientist and then general engineer. I'd say to have it going at cycle 50 definitely is doable, maybe even cycle 40.

Everything in this build is optimized in order to extract the most cold out of the ice as possible to pre-cool the water before the right-most two tiles in the freezer which are the only places the water will freeze. Having temp shift plates in the freezer have a negative effect on the power consumption and output.

The whole thing runs on a single 2KW power rail and uses 759W on average (measured on a massive battery bank) to produce 8kg/s water at about 2.4C.

Note, the airflow tiles are a vacuum to provide perfect insulation between the freezer and melter.

image.thumb.png.e821b784b56beab65201b8067f5e0bc1.png

Spoiler

Plumbing is pretty straightforward. Bilge pump on the right. Feeder pump on the left. And a standard aquatuner loop.

image.thumb.png.cc1949b56f0bd36a4ffdf6036ff1150d.png

Wiring is just connecting everything together, not much to it.

image.thumb.png.c295d81244094cbe75017948af965fac.png

Conveyors is where it gets slightly more interesting. A lot of time was spent optimizing this. Ice is picked up and from the loader runs through the water cooling it down slightly while heating up the ice a tiny amount. Then it runs on the left most side cooling down the feeding water while gaining a bit higher temperature itself. The ice will never melt while in one of the left most tiles. The maximum temperature it reaches is -1.2C but most of the time it sits around -1.4-1.5C before the door melter loop.

image.thumb.png.ccebc12deb3e2193b691dc585bf43fe3.png

Automation is where it gets interesting. The aquatuner loop is set to -8C. Effectively that makes the right most metal tiles in the freezer average -12C which is why the water freezes so fast  The tepidizer is set to on above 2C through a NOT from the temp sensor and via a 0.1s buffer from the pipe temp sensor meaning that the tepidizer will only turn on if the aquatuner is off and the water temperature reaches below 2C. Bilge pump is set to run above 1050kg. And the feeder pump runs when water pressure in the freezer is below 200kg and the water pressure in the melter is below 1200kg. The latter to avoid pressure build up.

image.thumb.png.7b09fc399cf517d072e727f3d310256c.png

Thanks to @mathmanican for providing the groundwork and proof of concept.

I should note that with more exotic materials this will reach 10kg/s output fairly easily at a higher power cost per kg/s of water but this build was made specifically in order to have minimal power and material costs so it could be rushed every game. Should you need more than the 8kg/s this will provide please refer to the original build.

Link to comment
Share on other sites

I am a bit unhappy with the game/current state..
image.thumb.png.f3473110f73762eadaee23d62b841f4a.png

I have very weird stuff happening too..
Only polluted water is venting in (runs through filter). I have now exactly 11.1 kg water there (only pw venting through), no blocked vent shows up and it turns into PW again..
No chance that 11.1 kg water got there accidentally..
WTF
image.thumb.png.01838c7a9652efb51f5b19a935532794.png
image.thumb.png.76e63e69227e13976fedb4610c6899ea.png

Link to comment
Share on other sites

7 minutes ago, Oozinator said:

I am a bit unhappy with the game/current state..
image.thumb.png.f3473110f73762eadaee23d62b841f4a.png

I have very weird stuff happening too..
Only polluted water is venting in (runs through filter). I have now exactly 11.1 kg water there (only pw venting through), no blocked vent shows up and it turns into PW again..
No chance that 11.1 kg water got there accidentally..
WTF
 

Are you absolutely sure there could have being regular water coming in somehow? I have a mechanical filter on all my permanent infinite liquid storages now to make sure there's no possibility of having the wrong liquid accidentally vent into it. Learned it the hard way :D

Link to comment
Share on other sites

30 minutes ago, Saturnus said:

Are you absolutely sure there could have being regular water coming in somehow? I have a mechanical filter on all my permanent infinite liquid storages now to make sure there's no possibility of having the wrong liquid accidentally vent into it. Learned it the hard way :D

Yes sure, but how ever should vent 11.1 kg other element, when one element per tile is rule?
The "reservoir" was running hundreds of cycles, without any (clear/pre)water in it.
I had other identic setups (use them for all sort of liquids). When it's pressurized and another element enters, vent gets blocked, had that in the past.
This thing is different.. Incoming pw passes the overpressurized water tile..

Link to comment
Share on other sites

2 minutes ago, Saturnus said:

It's 11.1kg. Vents don't overpressurize until you reach 1000kg.

mhh but how do the PW pass through this water tile, it's still working?
When i mix minimal other liquid into it -> get's blocked (waterfall mechanis breaks) for me..
I could always only store one liquid in such a "tank" it failed always when stuff mixed up..

Link to comment
Share on other sites

1 minute ago, Oozinator said:

mhh but how do the PW pass through this water tile, it's still working?
When i mix minimal other liquid into it get's blocked (waterfall mechanis breaks) for me..

It works because it's water over the vent. It's the lightest liquid so will stay over the vent regardless of which liquid you have in the storage. Had it been crude oil for example the crude oil would displace the PW and stop the vent.

Link to comment
Share on other sites

6 minutes ago, Saturnus said:

It works because it's water over the vent. It's the lightest liquid so will stay over the vent regardless of which liquid you have in the storage. Had it been crude oil for example the crude oil would displace the PW and stop the vent.

I do not understand why it failed the other times, but i trust you.
You win the knowledge game spock ^^

Link to comment
Share on other sites

Just now, Oozinator said:

I do not understand why it failed the other times, but i trust you.
You win the knowledge game spock ^^

3.14159265359 cheers for being a nerd :D 

(Btw, I remember pi by reciting this phrase: May I have a small container of coffee please sir Chris Bonington.)

Link to comment
Share on other sites

While doing my "mini"-geyser build above. I was puzzled why I couldn't get it above 8kg/s. I found the cause. The hydro sensor governing the bilge pump would cycle up and down violently when a piece of ice melted in the door next to it. From 1100 steady state to 200-300 momentarily and then back again. This was what caused why I didn't have full, or almost full output. Fixed it with a 10s buffer gate so now we're at 9.8kg/s output for about 960W average consumption. Really the output is 10kg/s most cycles but there's still an occasional hick up.

However, the increase in output cascaded a lot of new small issues that had to be fixed.

Anyway, the result is running smoothly now. Still no space age or otherwise exotic material used (including diamond). Instead of going through all the small changes one by one and why they were made, here's the save file so you can have a look yourself.

Timeline.sav

One thing I will mention as @mathmanican will probably like to know. The improved conveyor loop now makes sure that practically all ice is melted inside a door. That is what causes the doubling of the mass. It happens when the ice melt and is displaced from within a tile. If you watch it in the step function you can see the ice being ejected at the same time as the water is being ejected, effectively giving you double the mass as both happens at the same time. I'm very confident of that.

Link to comment
Share on other sites

So, let me get this straight. We can now create near 10kg/s of clean cool water, only costing us electricity. We can also directly convert any gas to any other gas we want. 

So if we siphon off 1kg/s to run an electrolizer and convert both the hydrogen and oxygen to natural gas, we have enough to run 11 natural gas generators assuming no hiccups in the electrolizer. 

I think @mathmanicanand @Saturnus may have finally found an exploit to beat the Borg Cube! 

Link to comment
Share on other sites

13 minutes ago, beowulf2010 said:

So, let me get this straight. We can now create near 10kg/s of clean cool water, only costing us electricity. We can also directly convert any gas to any other gas we want. 

So if we siphon off 1kg/s to run an electrolizer and convert both the hydrogen and oxygen to natural gas, we have enough to run 11 natural gas generators assuming no hiccups in the electrolizer. 

I think @mathmanicanand @Saturnus may have finally found an exploit to beat the Borg Cube! 

All credit should rightly go to @mathmanican. My contribution was minor in this regard.

However, no need to convert to natural gas. That just gives you the headache of having to convert the output of that. Better to use the semi-submerged electrolyzer exploit and get the full output, and then just convert as much oxygen into hydrogen as you need.

Note that this principle applies to any matter in the game that have both a liquid and a solid state. So you can have a polluted water "geyser" or a petroleum "geyser". They'll just be a little more tricky to build, and certainly will require exotic materials of some sort.

Oh, and note. It cost less than 1000W to generate 10kg/s water which when turned into hydrogen is enough for 80KW power. Crazy, right?

Link to comment
Share on other sites

45 minutes ago, Saturnus said:

Fixed it with a 10s buffer gate so now we're at 9.8kg/s output for about 960W average consumption. Really the output is 10kg/s most cycles but there's still an occasional hick up.

This parallels what I see with the high pressure build. I gathered data in debug to track it before posting it was shy of 10kg/s, and am glad you tracked down the reason. 

47 minutes ago, Saturnus said:

here's the save file so you can have a look yourself

I will be having some fun later today. Can't fire up ONI right now. 

15 minutes ago, beowulf2010 said:

So, let me get this straight. We can now create near 10kg/s of clean cool water, only costing us electricity. We can also directly convert any gas to any other gas we want. 

So if we siphon off 1kg/s to run an electrolizer and convert both the hydrogen and oxygen to natural gas, we have enough to run 11 natural gas generators assuming no hiccups in the electrolizer. 

It's actually WAY WORSE.

  1. Build yourself a 20kw power station with batteries and transformers (no input required, other than a tiny manual generator to kick start it). This can be accomplished very easily by cycle 10. Power is completely free. 
  2. Create the water (liquid) duplicator to give yourself a nice supply of 10kg/s water. Matter is free.
  3. Then convert the water directly to steam (A tepidizer will heat stuff up to 85C quite quickly. An exploited steam turbine will bring up the remaining 20C to boil it). Haven't built this yet, but just combines stuff I already know work. 
  4. Once you have steam, use the matter converter to get whatever gas you want at 10kg/s.  Skip the electrolyzer. This does require crude/petro in your matter converter. 
  5. If you want to cool it down rapidly, swap to chlorine and run 10kg/s chlorine past this water duplicator (the tepidizer inside won't fire as often).
  6. Then convert the cooled gas to whatever you want at 10kg/s. 
  7. Oh, maybe build a second liquid duplicator so you actually have water as well... Maybe.... 

Oxygen for 100 dupes? Water for your bristle berry plantation? Swimming pond for pacus? Carbon dioxide to feed your slickster farm that you never can keep fed? Natural gas for your generators (not sure why you need them...., just duplicate polluted water instead). Liquid phosphorus (gotta fertilize stuff, but dreckos are too slow - not really, but now I can just butcher them all). Liquid hydrogen for your rockts, I'm sure the dirt shortage can be addressed quite trivially with this. What are you missing? Just alchemy it. 

Haha.  Fun times. @Saturnus I'm glad you joined the fun. This would have not been even half as fun to work on without your input.  Thanks a ton for joining the ride. Now we just need to find out why it's not always 100% converting....  Still haven't cracked that. Nor have I cracked the 100x off gassing PO2 issues from a few weeks ago. 

My next post was suggested by @Oozinator, and I think it's definitely the plan.  Here's the title.

Exploits: Everything you need to know (or how to bug the **** out of ONI)

to be continued....

Link to comment
Share on other sites

1 hour ago, Saturnus said:

@mathmanican

Oh, and note. It cost less than 1000W to generate 10kg/s water which when turned into hydrogen is enough for 80KW power. Crazy, right?

The only reason I mentioned natural gas is for the 11th generator and the byproducts. If we're "wasting" 1kg/s, might as well recapture some of it as polluted water. But yeah, probably not necessary for "only" a 10% increase in the electricity output. 

Fun times. :D

Unrelated: Anyone figure out a good way to delete user tags on Android? Can't seem to delete Mathmanican's tag in the quote... 

Link to comment
Share on other sites

2 minutes ago, beowulf2010 said:

Anyone figure out a good way to delete user tags on Android? Can't seem to delete Mathmanican's tag in the quote..

I sometimes have similar issues. I have to delete lots around it.

I wonder if I can use this to exploit oni......

Link to comment
Share on other sites

4 hours ago, Saturnus said:

Instead of going through all the small changes one by one and why they were made, here's the save file so you can have a look yourself.

I had fun looking. I like the new conveyor setup of mostly conveyor bridges above the door. I'll adopt that for sure. I also liked some of the safeguards in automation. I did however notice that your ice builds up in the top area, so your current setup seems to not be able to keep up with the melting. In addition, the ice keeps dropping in temp, so making the melting even slower. After a few cycles I had 6t of ice in the top region. The feeder pump didn't stop once, but the bilge pump was not on always. Since you said it was running smoothing, I'm guessing that some of the heat transfer may depend on computer processing speed (my setup is not exactly optimal). You are definitely solidifying 10kg/s, but the melting speed was less on my CPU.

I built one of these in survival much earlier today, and noticed that if you surround it with a ring of airflow, followed by regular tile, it fits perfectly into a 4 tile per row setup (so plug and play into your base, wherever you want).  

Spoiler

Just prepping the area - doors will be replaced with airflow later.  Notice how the top and bottom of this fit perfectly in a 4 tile per row setup. 

5c8727ee7c71b_Screenshotfrom2019-03-1107-03-51.png.c2ff4eb710f4d8821f5438f25ddd92ed.png

The setup above has two rows of insulation, but next time I'll use an inner layer of airflow, and an outer layer of regular. Then I don't need to have an overflow, rather I can just keep the storage right here. This also has the advantage of maintaining the temp closer to exactly what you want (as the pressure builds), with very small fluctuations when the tepidizer fires.

Oh, fun side fact for those wanting to build this in survival. With the two liquid locks above (200kg PH2O and 200kg H2O), I had more than enough PW to fill the aquatuner loop after everything was built. I just popped a pump down to slurp up the airlocks (after mopping it all up and then clicking empty on the PW bottles).

In the future, I think I'm going to increase the width by 2 tiles, just so I can put another door in the bottom row to increase the melt speed. This would be more than enough to guarantee >10kg/s melt speed. I think the current 2 door setup in the base is perhaps just a tad too small to fully melt stuff.

 

Link to comment
Share on other sites

42 minutes ago, Saturnus said:

It's cubulicious.

It's a thing of beauty. :) After staring at it and salivating, luckily there was plenty of cool water to drink to re-hydrate. 

Unfortunately, when I loaded it, the aquatuner loop had stopped.  I believe this is because at some point you copied my incorrect version of the standard aquatuner loop (you had a different version earlier on that probably won't ever get stuck).  Recounting the pipe squares, I can see why this version is flawed - my bad. I learned something useful about perpetual loops from this post.

Or I just have a crappy computer. (and the crossed out stuff was true, and fixed)

 

Link to comment
Share on other sites

4 minutes ago, mathmanican said:

It's a thing of beauty. :) After staring at it and salivating, luckily there was plenty of cool water to drink to re-hydrate. 

Unfortunately, when I loaded it, the aquatuner loop had stopped.  I believe this is because at some point you copied my incorrect version of the standard aquatuner loop (you had a different version earlier on that probably won't ever get stuck).  Recounting the pipe squares, I can see why this version is flawed - my bad. I learned something useful about perpetual loops from this post.

I've save/reloaded, started and stopped this countless times and it's run 50 cycles with full 10kg/s output for all 50 cycles. And I ran it on both my laptop and workstation. I assure you, there nothing wrong with this set up. If you have any issues, the blame is squarely on your client or pc.

Link to comment
Share on other sites

2 minutes ago, mathmanican said:

I still have a crappy computer. No easy fix.

I'm just wondering why the hell it didn't stop at any of the other times. Especially when I loaded it remotely on my workstation to do a 50 cycle run on ultra.

Link to comment
Share on other sites

16 minutes ago, Saturnus said:

I'm just wondering why the hell it didn't stop at any of the other times. Especially when I loaded it remotely on my workstation to do a 50 cycle run on ultra.

I think the fix you implemented won't actually fix it.  The pipe lengths didn't change, just the positioning.

Theory (correct me if I'm wrong): The "aquatuner off" path needs to have the same or more pipe segments than the "aquatuner on" path, before reconnecting. The bridge insta-teleports, so we currently have 3 segments on the off path, and 4 segments for the on path. I think the following should fix this (add an extra pipe segment on the "off" line and then bridge in, swapping the 3:4 to 4:3 off:on pipe segment lengths). A fix to equalize the lengths would require a shutoff or valve, requires space, so scrap that. Here's the potential fix. 

5c88578fefd04_Screenshotfrom2019-03-1219-04-50.thumb.png.15d8710ea6c80c2210f9039893df887a.png

 I assume that's why you originally had the following (a 2:1 configuration, "off">"on"):

5c88578c56c0a_Screenshotfrom2019-03-1219-05-35.png.921c75852af887acf7da6742434aa00d.png

Why it didn't stop before? Good question. It should have. Mine should have long ago. Your save was the first time I saw it stop. The updated version above will often have a blob of PW sitting before the bridge, waiting it's turn, but there will always be several open spots for liquid to flow into, so it can't (shouldn't) stop. Once water stops on the "on"side, it's game over as the bridge can't push water forward to keep the flow going.

Please correct me if my thoughts here are wrong, but I think this is the issue. 

Oh, and @Oozinator, I'm waiting for the meme about us droning on about pipes.  Bring it on! Love em.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Please be aware that the content of this thread may be outdated and no longer applicable.

×
  • Create New...