Self powered phosphorite smelter 80

[bv55]

A self powered build with a 600 kg/cycle refined phosphorus output that fits in a 9*11 9*10  8*10rectangle. Requires supercoolant. 
Initial 9*11 version, for 9*10 version see upd3, for 8*10 see upd4

Spoiler

 

 

 

How it works:

It splits incoming phosphorite into 1kg pockets, then preheats it via counterflow by sending it through a tunnel with falling liquid phosphorus beads and an autosweeper in it, which speeds up heat exchange. After this phosphorite melts in a pool of liquid phosphorus that is kept hot by one of aquatuners, flows through a top tile of the phosphorus pool as its bottom is occupied by a drop of crude oil, falls down the tunnel, passes next to an aluminum block cooling down to 125 degrees and giving all the excess heat to a steam chamber, and then falls to the bottom, where it immediately solidifies. 

But cooling provided by an aquatuner that heats up a pool of phosphorus isn't enough for it, as solidifying 125C phosphorite requires around 60kdtu/s of cooling, and by liquifying preheated phosphorite we are getting only around 15kdtu/s, so a second aquatuner is needed to take care of it. To make sure no overload will occur it won't work when the first aquatuner is running. Also to keep the thing self powered it is linked to a battery, so it won't work unless battery is fully charged.

As heat exchange with the steam chamber happens only for a single tick the tile is filled with almost all kinds of bridges, pipes and wires imaginable to cramp as much mass as possible into it.

Steam turbine is self cooled and kept in a vacuum.

Automation: 

Spoiler

cycle sensor is set to 5 green/60 red, bottom temp sensor is set to below 250 with a buffer set to 15s, top temp sensor set to above 0. Conveyor meter set to one, battery is set to 50/90

Generally build can be shrunken up by cutting some corners and reducing steam chambers/cooling block size, but I really love rectangles
upd: I accidentally placed a radiant gas pipe under the bottom gas bridge, already replaced it with the insulated one in the game.
upd2: Also added a crude oil droplet on a tile next to bottom aluminum cooling block, as otherwise phosphorus solidified too quickly without cooling the conveyor loader. Phosphorus still instantly solidifies, but now loader exchanges heat with crude oil and stays at 40C

Spoiler

upd3: After some testing and an advice from @Zarquan I came up with a slightly smaller version that fits in a 9*10 rectangle. It uses a 5ton refined carbon tile as a heat battery with a buried thermosensor in it. The aluminum tile on top of melting chamber is needed to speed up phosphorus melting, as otherwise it became too slow after this update. Also added a petroleum loop that is used to check melting pots temperature,and a couple of extra bridges to even it out. Ran it for a view dozens of cycles and it seems to be stable.

Spoiler

 

upd4: Using another @Zarquan  idea I was able to shrink it even further to 8*10 by introducing a gas meter into it instead of the second aquatuner. Left liquid under the steam turbine is 10g of liquid naphtha, right liquid is sap. Insulated tiles/insulated pipes made out of ceramic. All the radiant things are aluminum. Everything with overheating temp is steel. Turbine itself is made from depleted uranium. Natural tile on the bottom is 5t of refined carbon, next to it is a droplet of supercoolant.
Automation: 

Spoiler

Top temp liquid sensor is set to below 250 with a buffer set to 15s, bottom sensor buried in a natural tile to above 0. Conveyor meter set to 1, Gas meter also set to 1. Battery set to 80/60. Cycle sensor set to 5 green/60 red

Spoiler

Latest yaml:

Spoiler

80 Phosphorus fixed.yaml

Edited April 12 by bv55

[KittenIsAGeek]

I've added phosphorite smelting to some of my metal volcano tamers using mid-game materials like steel and nectar for coolant, but haven't tried making a self-contained JUST for refining phosphorite.  

I like your hydrogen gas thermal transfer loops. 

[Zarquan]

Have you considered replacing the liquid bridge on the loop in the steam room with an aluminum conduction panel?  It works just like a bridge for your build, but is far more conductive.

What is the area up top with the supercoolant and metal tiles for?  A heat battery?

 

I would think this would likely be self powering without using supercoolant. 

The theory is that you get more energy from cooling the liquid phosphorus from ~250 C to ~125 C and using a steam turbine for that than you would heating up the phosphorite from ~40 C to ~250 C using an aquatuner on water.  I could be wrong though.

[bv55]

4 hours ago, Zarquan said:

Have you considered replacing the liquid bridge on the loop in the steam room with an aluminum conduction panel?  It works just like a bridge for your build, but is far more conductive.

I haven't thought about it, that's a good idea. But even with granite bridge it already instantly cools down to steam room temperature +1-3C, giving up almost all the stored extra heat energy, so it won't improve much. Although probably with it the extra hydrogen loop won't be needed
 

Spoiler

 

4 hours ago, Zarquan said:

What is the area up top with the supercoolant and metal tiles for?  A heat battery?

Yes, it is a heat storage that evens out the cooling loop temperature and measures where it stays on average, so an aquatuner stored in the steam room will have a better idea of when to turn on. It can probably be a couple of tiles smaller. I haven't experimented with its size too much in my final design. Removing it completely would be an issue though, as then the cooling block on the bottom won't even out the cooling loops temperature properly. Although with a supercoolant it shouldn't be that bad as it won't freeze.
 

4 hours ago, Zarquan said:

I would think this would likely be self powering without using supercoolant. 

I also think it can be (read post scriptums), but then the build will be at least two tiles taller for a couple of reasons. First reason is that incoming phosphorite doesn't get preheated properly because half of the time it gets through those two rails used to preheat it without touching a falling bead, so a tunnel will have to be at least two tiles taller than that, maybe even 4. Second is that a bigger cooling block will be needed to even out the cooling block temperature, as otherwise water will cause pipes to burst. Although if nectar is accessible it will prob work even with the current cooling block size.

P.S.

On second thought nope, it won't be. If my calculations are correct heating up 1kg of 30C phosphorite to melting temperature costs 32,250kdtu (31.5kdtu if it is at 40C), and cooling it down to 125 gives 96,212 kdtu back. So considering steam turbines 0.969 coefficient only 60w is left on cooling it to 45C and its transportation. And to cool it down to 45C would cost 126w with water and similar liquids, and if water is used as coolant in an aquatuner only half of this will be recovered, so it would cost around 60w. And in the end no energy will be left on phosphorus transportation

P.P.S.

On third thought aquatuner heating up phosphorus will give us some cooling, but it will cost double the energy of the one in the steam room. In the best case scenario 6w is required on conveyor loader to transport 1kg/s and lets assume that we properly min maxed sweeper and it also only costs 6w this will sum up to 12w. Even considering all the rounding involved and the best case scenario this will still be too much. So even maximally optimized phosphorite smelter that uses water as a coolant will still cost around 10-20w. And to get to this numbers its preheating tunnel will have to be much bigger.

And maximally optimized energy wise build running on supercoolant will only be able to output 40-45w. This build if hooked up to the grid will probably output 30w (based on battery overcharging from time to time), but I didn't measure it properly so I don't have the exact numbers. 
Edit: Did some testing, it outputs 25w because it was able to consistently overcharge the battery with 5 deodorizers hooked to it, and wasn't able to do it after 6th was introduced
 

Edited April 9 by bv55

[bv55]

The only way I can think of to make this thing power positive while running on water (outside of cheating by using 20 transformers in its steam room, adding a room with a wild plug slug in it etc) is to somehow incorporate a liquid uranium phase change based heat multiplying reactor in it. But even then I don't think it will work, and it will be ridiculous size wise

Edited April 9 by bv55

[Zarquan]

3 hours ago, bv55 said:

Yes, it is a heat storage that evens out the cooling loop temperature and measures where it stays on average, so an aquatuner stored in the steam room will have a better idea of when to turn on. It can probably be a couple of tiles smaller. I haven't experimented with its size too much in my final design. Removing it completely would be an issue though, as then the cooling block on the bottom won't even out the cooling loops temperature properly. Although with a supercoolant it shouldn't be that bad as it won't freeze.

I'm a big fan of natural refined carbon blocks for heat batteries.  You can make them to absurdly high masses using the automatic dispenser trick, where you use a dispenser to add refined carbon debris to an existing refined carbon tile as a buried object, then save and load to see it all merge, or just make them 4000 kg (equivalent to about 1670 kg water) by melting temp shift plates over them.  I use these natural refined carbon heat batteries all over the place because they can be incorporated into the walls and require minimal setup, especially if you have access to hot liquids. 

I generally make the initial tile by dropping 1 kg or 1 g (depending on whether I feel like setting up the shipping) on a hot plate until it melts.  Everything that will ever be behind that tile will be there.  I then either melt coal temp shift plates in the tile above it or build an automatic dispenser with it's nozzle in the block to add more mass.

I'd put one in place of the left metal tile in the phosphorus freezing chamber, so then you likely wouldn't need an external heat battery as long as you add all the heat conduction bridges (automation, maybe a conduction panel, conveyor, tempshift plate) and whatnot.  You can even put the thermosensor in before you make the natural tile.

Here's one in my phosphorus melter (which is heated through a hydrogen pipe).

Spoiler

This could let you push the whole phosphorus-melting area up a tile or two, letting you make the whole build a tile shorter.

3 hours ago, bv55 said:

And maximally optimized energy wise build running on supercoolant will only be able to ouput 40-45w. This build if hooked up to the grid will probably output 30w (based on battery overcharging from time to time), but I didn't measure it properly so I don't have the exact numbers.

There is one optimization you are missing, and that is using the output of the steam turbine to cool the phosphorus to about 98-99 C by further heating the output water before it gets vented back in to the steam room.  Basically, after you self-cool, you can run the output water through a small 2x3 heat exchanger that's just two metal tiles with a gap between them, then run 1000 g/s (using the 10% pipe exploit) through one tile and the phosphorus through the other.

I think you could ditch the second aquatuner by introducing a gas pipe loop to transfer heat from the phosphorus to the steam room when the phosphorus room gets too hot.  That would either shrink the build or let you add something else.  What might make sense is an aquatuner outside either chamber that transfers heat via gas pipes. 

Edited April 9 by Zarquan

[bv55]

44 minutes ago, Zarquan said:

I'm a big fan of natural refined carbon blocks for heat batteries. 

Thanks, this is quite interesting. Never thought about carbon tiles being that easy to make.

44 minutes ago, Zarquan said:

There is one optimization you are missing, and that is using the output of the steam turbine to cool the phosphorus to about 98-99 C by further heating the output water before it gets vented back in to the steam room.  Basically, after you self-cool, you can run the output water through a small 2x3 heat exchanger that's just two metal tiles with a gap between them, then run 1000 g/s (using the 10% pipe exploit) through one tile and the phosphorus through the other.

True, didn't thought of it. With it it may be barely possible, but I am not sure. And size wise build will be bigger. 

44 minutes ago, Zarquan said:

Which reminds me, I think you could ditch the second aquatuner by introducing a gas pipe loop to transfer heat from the phosphorus to the steam room when the phosphorus room gets too hot.  That would either shrink the build or let you add something else.  What might make sense is an aquatuner outside either chamber that transfers heat via gas pipes.

This is what my initial plan was. But it caused too much heat to leak into the steam room, which made the build inefficient. So in the end I came up with a double aquatuner solution.

Spoiler

another idea was to use wheezworts, but it needed at least three of them to work, and I wasn't able to fit more then one of them in there without increasing this boxes size

Edit:

44 minutes ago, Zarquan said:

This could let you push the whole phosphorus-melting area up a tile or two, letting you make the whole build a tile shorter.

it will help to cut a corner, but it won't make the build shorter. Wall on the right from the melting chamber is required for beads to form

Edited April 9 by bv55

[bv55]

2 hours ago, Zarquan said:

This could let you push the whole phosphorus-melting area up a tile or two, letting you make the whole build a tile shorter.

Using the carbon block advice I was able to shrink the build by 1 tile, but in a slightly different manner nvm, build needs some more testing. Seems like cooling block of just this size is a bit too small, and with it build runs into some issues. Will test it out, maybe it isn't the case but idk.
Edit: Found the problem, it was the conveyor rail which was now going on top of the carbon block

Spoiler

 

 

 

 

Edited April 9 by bv55

[bv55]

4 hours ago, Zarquan said:

This could let you push the whole phosphorus-melting area up a tile or two, letting you make the whole build a tile shorter.

Okay, now I have done it.

Edit: added a couple of bridges here and there, as otherwise new thermosensor was too inaccurate, which made the build inefficient

Spoiler

 

 

 

Edited April 9 by bv55

[Zarquan]

So, I created a version of your build with a few modifications I think could be beneficial.

First of all, I supplemented the self-cooling mechanism with phosphorite.  This has a few benefits, as it maintains the steam turbine at a lower temperature, meaning we can have a hotter steam room (with phosphorite incoming at 26.8 C).

I also ran the phosphorite through the steam room.  The best way to use heat in this build is to heat the phosphorite.  It is more efficient per DTU to heat the phosphorite than it is to run a steam turbine and use an aquatuner.

I removed the aquatuner within the steam room and replaced it with a hydrogen gas loop with a shutoff valve.  If the room gets too cold, it turns on the hydrogen loop.  I unfortunately do not, at this time, have a mechanism for detecting that the steam room is too hot, but I suspect I could add a pipe temperature sensor to one of the loops somewhere.

I created a new path for the conveyors because I didn't have room for the chute anymore.

I added automation to make it so that the aquatuner can't be on for more than 10 seconds at a time in a 50 second period to avoid what I call thrashing on the aquatuner, where the aquatuner keeps turning on and off due the battery being empty.  It's crude and could be better, but I'll leave that for a potential future attempt.

I could probably make it one tile narrower by moving the timer sensor and the conveyor meter up by 1 and putting in refined carbon + aluminum tile combos, then using a pipe loop to move the heat, but that is work for another attempt.

Because I made thee room one shorter, I had to move the steam turbine to prevent the phosphorus bead from breaking.  I tried putting 10 g naphtha with the original position, but I think it overheated the steam turbine.  Not entirely certain how to fix that with the smaller steam room, and I don't think it can be shorter and rectangular with this configuration without making the steam room smaller.

 

I will measure the net energy gain over 6 cycles using the debug tool on the selected 14 tiles, as that is all heat energy that can be converted in to steam power for the base. The difference is the total extra heat generated, which I believe to be a fair metric for the effectiveness of the build, though I do believe that a cooler steam room leads to higher efficiency, as it makes the aquatuner do less work on cooling the phosphorus.

Edited April 9 by Zarquan

[Zarquan]

This is starting to feel like a challenge build, which is fun.  How do you feel about entombed buildings that work (e.g. liquid valves, meter valves, surprisingly not liquid shutoffs.)

3 hours ago, bv55 said:

Okay, now I have done it.

Edit: added a couple of bridges here and there, as otherwise new thermosensor was too inaccurate, which made the build inefficient

  Hide contents

 

 

 

 

Nice, got it one narrower.  Are you sure you need the upper cold heat battery?  What if your aluminum tile was just a heat battery?

Edited April 9 by Zarquan

[Zarquan]

I decided to not take in to account the phosphorus, as it changes mass a lot.

It looks like my build generated an excess of 81.67 MDTU over 6 cycles, or 22.686 kDTU/s. 

[bv55]

12 hours ago, Zarquan said:

First of all, I supplemented the self-cooling mechanism with phosphorite.  This has a few benefits, as it maintains the steam turbine at a lower temperature, meaning we can have a hotter steam room (with phosphorite incoming at 26.8 C).

Love this idea. although I think it should be passed at 1 tile height, so it won't exchange heat with steam turbines foundation tiles

12 hours ago, Zarquan said:

I also ran the phosphorite through the steam room.  The best way to use heat in this build is to heat the phosphorite.  It is more efficient per DTU to heat the phosphorite than it is to run a steam turbine and use an aquatuner.

I thought about it, but I was afraid that it might cool the steam room too much under some unusual circumstances, so I dropped this idea. But I think in the end it was a mistake on my side.

12 hours ago, Zarquan said:

I created a new path for the conveyors because I didn't have room for the chute anymore.

Initially I tried to make a chuteless build, but then the empty conveyor belt bug got me. Looping around phosphorite will probably fix it, although I don't like the idea of passing the hot phosphorite through the whole thing again and again stealing energy from melting chamber

 

12 hours ago, Zarquan said:

I added automation to make it so that the aquatuner can't be on for more than 10 seconds at a time in a 50 second period to avoid what I call thrashing on the aquatuner, where the aquatuner keeps turning on and off due the battery being empty.  It's crude and could be better, but I'll leave that for a potential future attempt.

I am also trying to avoid it at all cost, but my solution is a bit different. I set up steam chambers aquatuner so it will run only when battery is 90% full, and only until it is down to 60%. And set melting chambers aquatuner so it will run for at least 15 seconds, which will drain around half of the battery. And as the system is energy positive and 60>50 it will always have enough power stored for it.

With a single aquatuner imo balance between aquatuners minimum uptime and the temperature range in which chamber will be between this heat ups will be more important then it is with two.

11 hours ago, Zarquan said:

This is starting to feel like a challenge build, which is fun.  How do you feel about entombed buildings that work (e.g. liquid valves, meter valves, surprisingly not liquid shutoffs.)

Nice, got it one narrower.  Are you sure you need the upper cold heat battery?  What if your aluminum tile was just a heat battery?

I also think builds are fun, as long as we aren't building a sour gas boiler design №19123113. I am fine with entombed stuff, as long as the buid is reasonable and possible outside of sandbox.
Well, I don't want to rely on a temp sensor on a pipe that will fluctuate by 5 degrees every several ticks. It will work with a single carbon tile with a thermosensor inside though, as it will measure the average temp which wouldn't fluctuate much. It will cause temperature of supercoolant coming into an aquatuner to differ greately though, but it shouldn't be that much of an issue as it won't differ by 500C, so it will be always ready to run.

8 hours ago, Zarquan said:

I decided to not take in to account the phosphorus, as it changes mass a lot.looks like my build generated an excess of 81.67 MDTU over 6 cycles, or 22.686 kDTU/s. 

So it is also around 25w, if I am not missing something

Edit:

So I tried it with a single carbon block for cooling, it doesn't work for me as the system becomes too unstable, and the melting chamber constantly spills its content with it. With the top cooling block system seems to be really stable and durable.
Edit2:
Yep, after trying again and again it gets energy negative after the second block removal, which eventually causes all kinds of things to happen, including chamber spillage and such. Why? I have no idea.

Edited April 9 by bv55

[bv55]

11 hours ago, Zarquan said:

This is starting to feel like a challenge build, which is fun.  How do you feel about entombed buildings that work (e.g. liquid valves, meter valves, surprisingly not liquid shutoffs.)

Nice, got it one narrower.  Are you sure you need the upper cold heat battery?  What if your aluminum tile was just a heat battery?

Found and fixed my issues, now it works without the top cooling block

7 hours ago, bv55 said:

I also think builds are fun, as long as we aren't building a sour gas boiler design №19123113.

I mean, I like the general idea behind sour gas boilers, but there are so many builds available online for all kinds of situations that it isn't fun anymore. And in terms of optimization - there are already all kinds of uber optimized builds available, so there isn't much to invent.

7 hours ago, bv55 said:

Borrowed

*Not borrowed, entombed

Edited April 9 by bv55

[bv55]

15 hours ago, Zarquan said:

So, I created a version of your build with a few modifications I think could be beneficial.

First of all, I supplemented the self-cooling mechanism with phosphorite.  This has a few benefits, as it maintains the steam turbine at a lower temperature, meaning we can have a hotter steam room (with phosphorite incoming at 26.8 C).

I also ran the phosphorite through the steam room.  The best way to use heat in this build is to heat the phosphorite.  It is more efficient per DTU to heat the phosphorite than it is to run a steam turbine and use an aquatuner.

I removed the aquatuner within the steam room and replaced it with a hydrogen gas loop with a shutoff valve.  If the room gets too cold, it turns on the hydrogen loop.  I unfortunately do not, at this time, have a mechanism for detecting that the steam room is too hot, but I suspect I could add a pipe temperature sensor to one of the loops somewhere.

I created a new path for the conveyors because I didn't have room for the chute anymore.

I added automation to make it so that the aquatuner can't be on for more than 10 seconds at a time in a 50 second period to avoid what I call thrashing on the aquatuner, where the aquatuner keeps turning on and off due the battery being empty.  It's crude and could be better, but I'll leave that for a potential future attempt.

I could probably make it one tile narrower by moving the timer sensor and the conveyor meter up by 1 and putting in refined carbon + aluminum tile combos, then using a pipe loop to move the heat, but that is work for another attempt.

Because I made thee room one shorter, I had to move the steam turbine to prevent the phosphorus bead from breaking.  I tried putting 10 g naphtha with the original position, but I think it overheated the steam turbine.  Not entirely certain how to fix that with the smaller steam room, and I don't think it can be shorter and rectangular with this configuration without making the steam room smaller.

 

 

I will measure the net energy gain over 6 cycles using the debug tool on the selected 14 tiles, as that is all heat energy that can be converted in to steam power for the base. The difference is the total extra heat generated, which I believe to be a fair metric for the effectiveness of the build, though I do believe that a cooler steam room leads to higher efficiency, as it makes the aquatuner do less work on cooling the phosphorus.

Thanks for the idea with the gas pipe, I was able to shorten the build using it. Solved the bead creating issue by placing a droplet of nafta on opposite side. Build will require some tweaking for efficency and durability, but I am pretty happy with it right now. The only thing that bothers me is that ceramic gas pipe with hydrogen tend to leak a lot of heat, but because they are burried in insulated tiles it isn't that much of an issue, although this means that the build will leak a decent amount of heat into pipes for the first 60+ cycles.

Edited April 9 by bv55

[Zarquan]

I attempted a build that tries to melt the phosphorite in the steam room and pump it.  While I do things like cool phosphorus with steam turbine output in normal games, these steam turbines don't run enough to provide meaningful cooling for the refined phosphorus, meaning that I won't get the benefit with a steam turbine dedicated to the build.  My steam turbines in normal games fulfill multiple functions, like also cooling volcano output or handling base cooling.

But, I did figure out the automation for the following:

while green, run for 10 seconds every 40 seconds where the 10 seconds are not interrupted by the signal turning red and the signal turning green promptly.

The main automation is at the top.  It is triggered by the thermosensor at the bottom right.  The thermosensor hits the "set" port on a memory module, which then sends a green signal to filter 1, filter 2, and one port on the XOR.  Filter 1 is the activation interval (40 seconds).  Filter 2 is set to the active time plus 1 (11 seconds).  Filter 2 means the XOR is one red and one green, so it outputs green and outputs to Filter 3, which is set to 1 second, which prevents green flashes from impacting the aquatuner.  Then, Filter 3 activates the aquatuner.

Also, look at this horrible liquid stack....  That crude oil in the bottom is 700 kg...

 

Just now, bv55 said:

Thanks for the idea with the gas pipe, I was able to shorten the build using it. Solved the bead creating issue by placing a droplet of nafta on opposite side. Build will require some tweaking for efficency and durability, but I am pretty happy with it right now. The only thing that bothers me is that ceramic gas pipe with hydrogen tend to leak a lot of heat, but because they are burried in insulated tiles it isn't that much of an issue, although this means that the build will leak a decent amount of heat into pipes for the first 60+ cycles.

Hey, I actually did that trick with the liquid in my build, but I determined (maybe falsely) that the phosphorus touching the naphtha was bad.  But maybe I overestimated how bad it is. 

Edited April 9 by Zarquan

[bv55]

37 minutes ago, Zarquan said:

Hey, I actually did that trick with the liquid in my build, but I determined (maybe falsely) that the phosphorus touching the naphtha was bad.  But maybe I overestimated how bad it is. 

It drains around 5-7C from the falling phosphorus.
There is a pudle of sap next to it, which  also is somewhat of an insulator, so 10g naphta and the sap don't exchange much heat. Also I switched steam turbines material to depleted uranium.

37 minutes ago, Zarquan said:

Also, look at this horrible liquid stack....  That crude oil in the bottom is 700 

So it is a pool of low mass crude oil, on which stays a pool of low mass petroleum, on which stays a pool of low mass liquid phosphorus. An attempt to arrange something like this in survival would be fun. And seems like it was done to cool the phosphorus coming from the steam chamber, and probably preheat an incoming phosphorite.

Also I like your aquatuner, it reminds me of a Hogwarts logo😄

Edited April 9 by bv55

[Zarquan]

36 minutes ago, bv55 said:

Also I like your aquatuner, it reminds me of a Hogwarts logo😄

I was actually referring to the aquatuner stack as the horrible one.  But the other stack is also annoying to set up.  Anything that relies on molten sulfur is likely to be annoying.

Edited April 9 by Zarquan

[Zarquan]

1 hour ago, bv55 said:

Thanks for the idea with the gas pipe, I was able to shorten the build using it. Solved the bead creating issue by placing a droplet of nafta on opposite side. Build will require some tweaking for efficency and durability, but I am pretty happy with it right now. The only thing that bothers me is that ceramic gas pipe with hydrogen tend to leak a lot of heat, but because they are burried in insulated tiles it isn't that much of an issue, although this means that the build will leak a decent amount of heat into pipes for the first 60+ cycles.

I think this is about as small as you can get this while still having all this equipment built in to it.  I don't think it can be built any smaller (and rectangular) without removing key components.

Edited April 9 by Zarquan

[bv55]

4 hours ago, Zarquan said:

I think this is about as small as you can get this while still having all this equipment built in to it.  I don't think it can be built any smaller (and rectangular) without removing key components.

I also think so, at least in my variation. I am not sure about yours, as I don't know exactly what are all those metal tiles and natural tiles doing.
Tested it out - and it stabilized after awhile. It isn't as efficient as it could be, but it is pretty reliable. But it now takes quite some time to stabilize, as gas pipes continuously drain the heat for 100+ cycles.

Spoiler

 

 

 

[Zarquan]

On 4/9/2026 at 9:42 PM, bv55 said:

I am not sure about yours, as I don't know exactly what are all those metal tiles and natural tiles doing.

Those are a heat exchanger with the coolant coming out of the steam turbine.  The issue is that the steam turbine doesn't run enough to make much of a difference, so it doesn't work.  The natural tiles are two heat batteries, one for the steam room and one for the cold brick.

It would work if the steam turbine was doing more things, like cooling a volcano, so that it was active more often.  Basically, if this were attached to a steam room that is capable of supporting an always-on steam turbine to constantly send 95 C water through those pipes at 1000 g/s, this build may be more efficient than yours.  But it doesn't, and the only way to get that is to not make it a self-contained build, which (1) violates the implied rules of the challenge, (2) becomes annoying to measure the effectiveness and performance of, so the build fails to function, and (3) would be much larger due to the increased functions it would have to perform.  Although, you could say that this build uses a percent of a steam turbine operating at 200 C (say, 5% as a guess).

Maybe if I blocked another input....then the steam turbine would generate less power and run more.

Oh, I never actually run steam turbines with blocked inputs.  I guess they don't take 2 kg at that point, so I can do a little redesigning....

Edited April 11 by Zarquan

[Zarquan]

Ok, I think I've managed it.  A new contender!

The general idea is that I use the phosphorite and water to cool the turbine, then I counterflow through the three metal blocks, with the output from the steam turbine also counterflowing through the last two. It is the same size as your current best.

There are several automations in this that are of interest.

Spoiler

Every time the conveyor meter finishes it's meter value, it resets and ticks up the counter.  When the counter hits 10, it activates the liquid vent for about a second, letting one 10 kg packet of liquid phosphorus out.

This is the aquatuner logic that I described before.  Basically, when there is a green signal, to the Set port, it will continually activate the aquatuner for, in this case, 10 seconds every 40 seconds, without interrupting the 10 second activation time if the signal turns green.

Power extraction:  If the thermo-sensor in my heat battery/shelf reads over 265 C AND neither the aquatuner or steam turbine have been active for 10 seconds, the power transformer turns on.  It is vital that this power actually be removed, so you might want some kind of system to ensure that the power is used.  Perhaps the steam turbine just turns on if a smart battery on the other side activates, wasting power.  I will count every second the power transformer runs over a period of time.

 

 

Next, the plumbing:

Spoiler

 

There are three major things happening here.  First, phosphorus is pumped out of the room, through the three counterflow areas, then out the vent,  This only happens after 10 packets of phosphorite come through.

The steam turbine's output goes around in a loop around the steam turbine, through a liquid valve set to 900 g/s, and to the vent.  This is to prevent damage for when the steam turbine decides to spontaneously emit a packet over 1 kg.  The valve is buried in glass aerogel (0.1 g).

The aquatuner's loop is a little bit janky, but it cools the cold block the phosphorus freezes on.  It is stop and go when the aquatuner is off, but it flows smoothly when the aquatuner is on.

There is also a small loop cooling my power transformer and small battery.

Now for the shipping:

Spoiler

The phosphorite enters the system.  Here, it is 10 kg/packet, but in reality it would be 20 kg/packet, but I can't imagine that changing anything.  The 10 kg packets move once every 10 seconds.  They first go through the turbine room, then the first two layers of the counterflow before being metered off to 1000 g/s.  It then flows through the last layer and in to a conveyor loop in the melting/steam room.  This system, combined with the automation, prevents phosphorite from accumulating in the room if it gets too cold and prevents the liquid phosphorus from being all pumped away.

The refined phosphorus output is pretty basic.  The autosweeper is set to be active for 0.02 cycles every 1.6 cycles to save power.

The phosphorus leaves the vent at around 140 C.  If this system was being used to process more heat, which could be fed in using a liquid or gas pipe loop, it would lower the temperature of the phosphorus because the steam turbine would run more.  This would increase the energy output of the phosphorus melter by more than the heat added, so such an addition would be a strong positive, though you would almost certainly need to cool your steam turbine with the aquatuner if you did that.

I will be back later with approximate power output number over 10 cycles.

All natural blocks are 4000 kg refined carbon.

Liquid material overlay:

Spoiler

The liquids around the aquatuner are 700 kg crude oil (for thermal mass connected directly to the aquatuner), 600 g petroleum (so it flows over the liquid vent and blocks an input), then 10 g sulfur and 10 g naphtha.

 

 

Edited April 13 by Zarquan

[Zarquan]

My automation was bugged, but over 6,000 seconds, my automation made 5,096 checks (which was supposed to be 6,000 checks) and it detected that the transformer was in use 75 of those checks.  Using that, I estimated that measured that my build generates an excess of 14.717 W.  I won't show that automation because it didn't work well, but it's clock was a timer sensor set to 0.1 green and 0.9 red, which should be 1 second but isn't. 

I did assume that the transformer was on for each measurement and that they were evenly spaced.

EDIT:  I made adjustments to my automation.  Now, I check every 2 ticks (0.4 seconds, setting a timer sensor to 0.2 green 0.2 red, and it seems accurate) to see if the transformer is on and use an array of signal counters to count how many ticks have gone by and how many of them the transformer was on for.  And the transformer was constantly pulling 1kW while active. 

I had another overnight experiment, so I ran this new one overnight too (about 166 cycles).  Over 250401 0.4 second intervals, the transformer was on for 6516 of them.  That means that I have an estimated upper bound output of 1000 W * 6516/250401 = 26 W.  I think this upper bound is pretty accurate, as in all likelihood there was an equal number of times that the transformer turned on in the red and turned off before the two ticks were over.

Edited April 14 by Zarquan