The "Breather" Natural Gas Refinery - A modular & "reasonably compact" 10kg/s boiler.

[DaHodge]

This has been quite the moving target - when I started this build I'd had a single colony about 1000 days in and was just looking to abuse tempshift plates to do single-tile phase changes (Quite useful, if not very energy-efficient without significant design consideration). Over the past 3 months, I've seen new boilers pop up on a weekly basis, continually challenging me to refine my design. I've drawn lots of inspiration from several of them, from NurdRage's OSHA Hater, to Ainsley4ever's Super Sour Gas Boiler. Then mathmanican's latest, frustratingly-excellent design popped up, claiming the same feat I was hoping to claim here (single aquatuner 10kg/s boiler with some headroom).

While this design might not be quite as cooling-efficient, I think there are a few other key areas it has merit and I hope this particular configuration can be some inspiration to others, even if pretty much every piece of it has been done before

So without further adieu:

The "Breather" Refinery:

 

Spoiler

Additional Views

 

Let's not forget the bottom of the 62-tile deep boiler pump shaft:

And with the magma boiler added:

Key talking points:

• Low usage of space materials (Need approximately 200-300 super coolant, that's it!)
• Very high heating efficiency (Roughly 20-25 kg of crude processed per 1 kg of magma when using magma as the heating element - ~14-18 kg of nat gas)
• Very high cooling efficiency (Runs 10 kg/s on a single aquatuner)
• Fairly well separated & most components are modular (with some design impacts)
  • The simple petroleum boiler element can be hooked up to any heat provider, it just needs to be kept above 548 C at all times (for my tests I hooked it up to a "volcano" magma heat sink.
  • The sulfur "cold-sink" can be expanded or rotated if necessary
  • The temperature exchangers can be lengthened
  • The boiler pump tower's length can be adjusted (the length affects the efficiency of both the cooling & heating elements, however!)
  •  
• Pressure is not a concern with the chained door pump - the highest pressure in this configuration hovers around 100kg/tile at the end of the door pump itself.
• Simple to start up & stop.
  • To prime the system, just need to run the aquatuner until the coolant reaches below -200C (or initially fill it with -250C supercoolant for best results).
  • The automation ensures the system will not overpressure the cooler or flood the boiler.

Why "Breather"?

Some nonsense about how this much natural gas "breathes" life into a colony. Or it's just that the way the door pump works causes some pressure variations even when the overall refinery has reached an equilibrium. You choose

Operational Details:

Main Refinery Flow

The Boiler Pump:

Crude oil enters from the pipe on the left. It is split up into two packets, 5kg each - this avoids losing bead mass due to flaking. The tower is a bead pump heat exchanger (as featured in many other boilers). The base is a simple steam heat sink (best way I've found to keep temperatures within a reasonable range, steam has such an amazing thermal profile) that must be kept above 548C, ideally as close to that as possible to ensure maximum efficiency (probably could go down to 546C but any lower seems to cause buildup and 2C of leeway doesn't lose us much efficiency here).

The Heat Exchanger:

Upon reaching the top of the pump, the sour gas cools to somewhere around 180-190C. Then begins the door pump into the heat exchanger. The first exchanger is with the sulfur "cold battery". This is yet another steam heat sink with the output sulfur running through it (until it reaches 0C). There are three outputs here - could likely make do with one or two but there was a good amount of space here. After passing across the top, it reaches the natural/sour gas exchanger. This is pretty standard - a bunch of conveyor bridges and alternating diamond tiles that help to balance the temperature between the two, making use of the fact that hot gas rises & cold gas sinks. The door pump ends before the bottom for two reasons. Firstly, I wanted to allow some buffer space to prevent blowing out the methane cooler with a huge wave of pressurized sour gas, and also to give enough room to allow access into the turbine room for maintenance.

After rounding through the heat exchangers, the sour gas near the bottom of the exchanger should ideally reach below -100C. Occasionally this will be higher as the refinery reaches equilibrium, and it will continually vary due to the injection of pressure from the pump, but as long as it can reach -90C-100C here the cooler will continually function.

The One-Tile™ Cooler:

Now we hit the One-Tile™ cooler. This is the first thing I came up with when designing a Sour Gas refinery and something that's always worked quite well throughout my various tests (you can see many of them in my save file). This beauty relies on a tempshift plate covered in supercoolant (or hydrogen, haven't tested yet) surrounded by supercooled diamond tiles & kept cold by a supercoolant loop. The top-right (or top-left in other configurations) corner is an airflow tile where the sour gas can flow into. The beauty here is that all that cooling power goes into that one tile, and when it chills to methane it is pushed up into the tile above, where the diamond tile & conveyor bridge exchange heat with the sour gas, heating the methane into natural gas.

This remains stable because natural gas is lighter than sour gas and will continue to float upwards from the exchanger tile, while sour gas will remain below. The automation here is an emergency shutoff - if sour gas breaches the cooler (usually due to an overpressure causing the cooler to stop working) the boiler will pause until the system has normalized.

On to the Ouptut:

From here, the natural gas exchanges heat with sour gas, reaching a temperature range of 100-125C. Pretty hot, but not as hot as the "OSHA Hater" - 125C is quite manageable and could easily be cooled further or pumped directly into Natural Gas Generators. The gas sensors & door at the top help prevent sour gas from entering the pump chamber - if either tile is not natural gas the door closes (it will start closed before the refinery begins operation).

Peripheral Systems

Sulfur Reclamation:

Maybe it's an exploit, but it's too convenient. This build uses an autosweeper w/ a missing corner tile to extract sulfur from the cooling chamber. This is deposited into a two-stage heat exchanger, the first stage being used to cool Sour Gas, the second used to cool the steam turbine room before outputting dupe-accessible sulfur into the turbine room. The liquid pump is there to extract any sulfur that overheated, generally due to small packets that quickly react to the 145+C temps of the sulfur exchanger. The automation is set to liquid sulfur, this prevents it from accidentally pumping out water before the exchanger has warmed up to standard operating temps (water and liquid sulfur can't exist side-by-side as if the room were cold enough for water it'd cool any sulfur back to solid form).

Breather™ Pump:

The pump system is a fairly straightforward 4-stage repeating door pump. I tried building this with standard automation wires and realized how much simpler it would be with a ribbon. The timer is set to 12on/8off and all of the delay/buffer gates are 5s each. The doors are Gold Amalgam - this leads to the lowest amount of conductivity between doors (which is already a low concern due to the ~1s overlap where neighboring doors are both closed). All dynamic automation is kept in dupe-accessible areas to allow for maintenance/replacement if necessary. Honestly, as long as there isn't a catastrophic failure (haven't encountered one in the ~100 cycles this thing has run so far) it should be possible to fully evacuate the system to be serviced by dupes.

It may be desirable to expand on the automation here to allow for full-open or full-shutoff of the door pump - given the ribbon automation system it should be fairly straightforward.

Magma Heat Sink

The magma heat sink I include as an example heat provider for the sour gas boiler is designed to maximize heat extraction of magma/igneous down to 650C. Other designs used magma to preheat crude, but I found that the boiler-pump exchanger was far more efficient for the entire process. The 650C igneous output could be used to generate supplementary energy to power the build or for some other nefarious purpose. My favorite part of the heat sink is the petroleum bead cooler, which is an improvement on the other vacuum-based igneous extractors. Rather than submerging them in a puddle of petroleum to facilitate heat exchange (which is risky as the autosweeper can drop materials upon being disabled or upon save/load) this system drops a few 5kg packets of cold petroleum over the autosweeper & conveyor loader, which fall through the mesh tiles & get pumped back into the system to be chilled. I didn't go through the trouble of creating the full loop, just dumped some cold petrol there, but this system is fairly efficient and avoids the big pitfall of superheated steam/sour gas. Even if igneous is dropped on the floor, it would take several seconds to heat up the petroleum to the point where it'd boil into sour gas (I tested this - it took about 3 seconds for a 5kg bead of petroleum to boil with 1200C igneous). These would only touch for a single tick!

Next Steps:

I'm likely going to regret this, but I'm going to attempt building this in survival. Though it is fairly extensive, I have tried to design it with survival in mind - the most complex parts are fairly small & condensed in the center (thus can be built first) whereas the outer sections are fairly straightforward (with the exception of the automation).

I hope this serves as some food for thought - while I was hoping to provide some net-new material to chew on I think this ended up just being a mashup & refinement of several other techniques. I don't think I've seen the single-tile cooler elsewhere, though

If anyone has any input - optimization ideas, flaws, or ways to make this more dupe-buildable/accessible, I'd love to hear it!

Save File:

The boiler featured here is the one just to the right of the dupe haven, but I kept many of my incremental designs (I'm sentimental and they were good for reference, though I probably deleted about half of them as I went).

UltraBoiler extra.sav

[mathmanican]

I love the one-tile cooler. Nice use of ONI physics.

[Garueri]

This is nuts, awesome job sir.

[kbn]

Although not a post on this forum, I have previously created an natural gas production system using a mechanism similar to this One-tile cooler.
(However, since it does not use as many techniques as OP, it is not very efficient. natural gas production is about 500 g/s)

 

Spoiler

 

There is an interesting technique I was using there.
Try adding a solid tile directly above the Airflow Tile of the One-tile cooler to completely separate the sour gas and natural gas.
Even if split, liquid methane should jump over the tiles and appear in the natural gas room.
This is a technique that uses the specification that when the gas becomes a liquid in the Airflow Tile, the solid tile above it is ignored and it jumps into the space beyond it.

Spoiler

I think it's already working fine without this technique, but with it you can completely eliminate the slight possibility of sour gas intrusion.
 

[suxkar]

4 minutes ago, kbn said:

 

  Hide contents

 

  Hide contents

I think it's already working fine without this technique, but with it you can completely eliminate the slight possibility of sour gas intrusion.
 

Lol

This is madness. I had no idea teleportation could be abused this much. The amount of possibilities

[TheMule]

Quick question... instead of making oil enter from the right, how about having the pipe come from below, and go through the steam room with some radiant pipes to soak up some of heat generated by the AT? Some kind of pre-heating.

There's room for 4 segments of thermium/aluminum radiant pipe then oil can just follow the insulated wall up to the same valve. There are 3 intakes open so the AT could run only if the temp reaches 270, but if you block another one you can easily reach 300-310C, enough to protect the steel AT. A temp sensor activates the turbine only above that, but at that temp, oil could drain enough heat from the steam for the turbine never being activated at all.

Would that alter the balance of the heat exchanger in a counter-productive way tho?

[DaHodge]

10 hours ago, kbn said:

Although not a post on this forum, I have previously created an natural gas production system using a mechanism similar to this One-tile cooler.
(However, since it does not use as many techniques as OP, it is not very efficient. natural gas production is about 500 g/s)

I love your refinery!! My original goal was to make a minimal-size refinery with no space materials but I saw too many high-throughput ones and found optimizing that to be a distracting challenge. Seeing yours makes me want to go back to that, though! Honestly, 10kg/s is excessive until late game when you likely have access to space materials anyways.

9 hours ago, kbn said:

Try adding a solid tile directly above the Airflow Tile of the One-tile cooler to completely separate the sour gas and natural gas.

That's almost unbelievable how effective that is - I'll have to give that a shot  The fact that it can teleport to an arbitrary height seems super useful. That looks almost related to the issue being discussed elsewhere where an automatic dispenser can create an infinite column of some material... this is just the gas form of that? XD

I think with this particular configuration it'll need some redesign - I'll have to expand the exchanger by 1 tile to keep the current temp balance, and the automation to prevent it from going overpressure will need to be reworked. I really wish atmo sensors could detect more than just 20kg... when you're processing 10kg the neighboring tile ends up being over 2x pressure x_x

Very exciting though - this this opens up some completely different possibilities on how to structure the exchanger.

 

2 hours ago, TheMule said:

Quick question... instead of making oil enter from the right, how about having the pipe come from below, and go through the steam room with some radiant pipes to soak up some of heat generated by the AT? Some kind of pre-heating.

...

Would that alter the balance of the heat exchanger in a counter-productive way tho?

It's a good idea for reducing the area taken by the build, but it actually does affect the temperature balance of the heat exchanger. This build is designed to maximize the temp exchange between crude oil & sour gas - the output sour gas temp of the bead tower is around 185C when the input is 100C. Having a temp below 200C is what makes the temp exchangers work so well without being even longer - they manage to get the sour gas temp down to -95C to -105C before hitting the freezer tile.

It's kinda funny - while most of the other boilers rely on to preheating crude *as much as possible* beforehand, this build actually benefits from colder crude input! Granted, it'd probably slightly reduce the boiler efficiency, but it wouldn't be as significant as the effect on the cooling.

The save actually has the (disabled) remnants of my previous attempts - most of them actually do have crude oil preheaters (either by using sour gas, a magma heat sink, or some other mechanism). One challenge they all suffered from was trying to reliably preheat as high as possible without bursting pipes. I really like your idea of using the heat from the cooling AT to preheat - could definitely build a compelling design around that  Removing the need for the turbine would be a huge space saver, but only if you can still get by with one AT

[TheMule]

51 minutes ago, DaHodge said:

Removing the need for the turbine would be a huge space saver, but only if you can still get by with one AT

I'm not sure removing the turbine would be entirely possible... the point is not having it run most of the time. Now, I'd admit the point is moot. The ST recovers most of the energy used by the AT anyway. More than 90%. So it's already efficient.

And removing it might not be possible as it could be needed to address heat unbalance during start up / shut down, for example.

But the idea was to use the heat from the AT to reduce the magma requirement. Technically, if you build it out of thermium, you could use it as the main heat source.

I'm not sure I can explain what's in my mind, but... well consider inputs and outputs in your system. You have a heat source (e.g. magma), crude oil in, nat gat + sulfur out, and a (huge) heat deletion thing, the steam turbine. All the heat deleted by the turbine must come from somewhere, and that's magma (let's ignore small SHC differences here). Reduce the heat deleted (re routing it inside the system) and you can reduce the heat in. The trick is also to reduce heat out as hot nat gas (but if you output 0C sulfur, you're way below the temp of incoming crude, so the extra heat must go somewhere).

[DaHodge]

2 hours ago, TheMule said:

use the heat from the AT to reduce the magma requirement

I see what you mean now - the preheating was just an example of how it could be used. So if we pumped the heat from the AT all the way down to the boiling chamber (initial thought is to just have an infinite hydrogen loop running between the two, not sure if it'd provide enough heat transfer), it could potentially *replace* the magma altogether. Given that the AT runs essentially 100% of the time when running at full capacity, I *really* like this idea. Depending on how the numbers work out, you either get rid of the turbine or the magma heater altogether (if you generate extra heat, dump the magma, if you need more heat, dump the turbine).

We know how many DTU/s this needs:
~500-700g of magma per second, reduced from original temp (1727 or so) down to 650C = ~750kDTU/s
A supercoolant 'tuner produces over a million DTU/s. Seems like the magma goes away

Time to give this a try!

[mathmanican]

FYI.  Everyone who has gasses horizontally next to each other, with low mass cold next to high mass hot, will need to redesign when the temp swap bug gets fixed. You have this in your sour gas cooling area as well as next to the one tile cooler. Every time you open a door, the bug has a chance to fire. 

My spiral boiler abuses thus bug a lot as well (didn't even realize it was an issue at the time). It's really hard to avoid. It's the same bug that hurts steam turbines.

Spoiler

 

 

[DaHodge]

24 minutes ago, mathmanican said:

FYI.  Everyone who has gasses horizontally next to each other, with low mass cold next to high mass hot, will need to redesign when the temp swap bug gets fixed. You have this in your sour gas cooling area as well as next to the one tile cooler. Every time you open a door, the bug has a chance to fire. 

Well this is quite unfortunate XD The temperature differences aren't that large but given their volume it's a pretty significant DTU loss. I was wondering why the temperature gradient was so significant across the top of the door pump area!

I do think this might still be fine, though - I'll probably have to make the Natural/Sour gas exchanger a bit taller to compensate (with more output heat).

In other news, I created a quick mod to the design as recommended by TheMule

This works quite well. Steam Turbine activates when the steam chamber is over 660, over about 3 cycles the boiler steam room has been steady at 561C (a little too high, actually!). So add 1200kg of Thermium to the requirements and this is now a thermally-closed system... almost. Given that it outputs natural gas at an average of 120C and sulfur at an average of 40C, it comes out to a total absolute output of 647kDTU/s, with 100C providing 630kDTU/s. So about 47kDTU/s extra - not too bad  Just add 5.5KW of power!

We'll see how this all shakes out when the temp-swap bug is fixed... I can see that having a decently large impact on the cooling section as each tile to the left decreases in temperature

[DaHodge]

@mathmanican I may have found a solution for the temp bug (in the cooling chamber). Conveniently using the liquid-teleport that @kbn mentioned:

The idea being that there is no temperature exchange happening on the horizontal tiles and as cold stays low, the cooled tile should not exchange energy with the tile above. Aaaand the teleport conveniently sends the methane right up into the top chamber. Pretty slick trick, kbn!

Could also remove temp exchange from other horizontal components. By changing the overall footprint to be taller, the same functionality could be maintained.

Of course, none of the adjustments will be necessary when the bug is fixed

I'm almost certain it'll need to offload more heat into the natural gas to maintain cooling with a single AT though.

[mathmanican]

2 hours ago, DaHodge said:

I may have found a solution

Definitely reduces it. You still have those three bottom tiles by the cooler to deal with. And getting the sour gas to the top could be tricky. 

I spent 2 hours today working on exploiting the flaking reset temp bug but could not get a good solution that avoided mass deletion with two gasses or temp swapping with just steam. I can generate consistently over 2000W with a single tuner, but temp swapping and mass deletion are my enemies (bugs fighting bugs). Tomorrow I'll see what I can do with sulfer and aluminium.

[Prince Mandor]

4 hours ago, DaHodge said:

... it outputs natural gas at an average of 120C ...

What happens when this gas porcessed by nat.gas generators? Is it produce polluted water, or steam?

If it produce steam, is it possible to heat gas a bit more, use steel pumps and generators, and feed resulting steam to turbine?

[TheMule]

7 hours ago, DaHodge said:

see what you mean now - the preheating was just an example of how it could be used.

Yes, But if you want to keep a steel AT, I can't think of anything else than preheating oil. Ideally you could try and inject heat in the middle of the exchanger (so that oil counterflows sour gas at first, then soaks heat from the steam, then sour gas again) but I have the feeling it's going to mess with the exchanger balance anyway, not matter where you inject extra heat, sour gas is going to come out of the exchanger hotter. Well, it's going to reach the boiling chamber hotter as well, reducing magma consumption. Like I wrote there's a gain but it's on paper (or in bits making up a forum comment), not easy to translate that in reality (or in bits making up a game on PC).

A thermium AT is a different story entirely.

Nothing new under the sun. That's how heat works, there's "good" heat and "bad" heat, depending on the application. Heat in 95C water is "bad" heat, meaning you can't extract it for anything useful (*) w/o spending energy, while heat in 200C steam is very useful. Here heat in 320C steam (steel AT) isn't that useful, heat in 600C steam (thermium AT) is.

(*) Well, you can use it to heat up -15C p water from a CSG, or counterflow other types of water for boiling. It's not entierly useless heat. It's just very situational while for steam >125C it's kind of obvious what to do with it.

[DaHodge]

36 minutes ago, TheMule said:

(*) Well, you can use it to heat up -15C p water from a CSG, or counterflow other types of water for boiling. It's not entierly useless heat. It's just very situational while for steam >125C it's kind of obvious what to do with it.

And that's the beauty of doing this in survival, you'll always find some use in a colony for the seemingly useless & annoying byproduct

[kbn]

15 hours ago, DaHodge said:

That looks almost related to the issue being discussed elsewhere where an automatic dispenser can create an infinite column of some material... this is just the gas form of that? XD

There is no basis, but it may be.
ONI seems to have the habit of teleporting upwards when tile generation is blocked for certain patterns.

Save & Load was required to generate a tower of tiles with a dispenser, but there are also cases where a tower is generated in real time.
As far as I know, there are two. A tile created by Shove Vole near the wall and a tile created by the temperature change of the rubble on the conveyor.

Spoiler

Since the conveyor can be intentionally triggered, it can be used to dynamically add or remove room effects by automation if used properly.
Perhaps something interesting can be done about the operation of the ranch. (I don't have any useful ideas yet)

[mathmanican]

1 hour ago, kbn said:

ONI seems to have the habit of teleporting upwards

I have long wished to crack the "Dupe" teleporting phenomenon. Upwards movement could then be done via teleportation, and downwards movement by firepoles. It would be hilarious to walk to a spot in the oil biome, and then on the next tick be up in space. Haven't found a way to achieve it yet.  

[kbn]

58 minutes ago, mathmanican said:

I have long wished to crack the "Dupe" teleporting phenomenon. Upwards movement could then be done via teleportation, and downwards movement by firepoles. It would be hilarious to walk to a spot in the oil biome, and then on the next tick be up in space. Haven't found a way to achieve it yet.  

It's not particularly useful, but you can teleport the wall from left to right...

Spoiler

 

 

Since I was interested, I searched for past Dupe teleport cases.
I've learned that there are many cases, from those that are fixed as bugs to those that are not.
I'm surprised that more Dupes were teleporting than expected. Certainly it would be very interesting if the method of teleportation could be elucidated.

[OxCD]

On 6/30/2020 at 2:26 PM, kbn said:

 

  Hide contents

 

This one, with the voles, is just annoying as hell. The simpliest ranch you have with voles, and many regolith go out of the ranch with this.

 

On 6/29/2020 at 12:10 PM, kbn said:

 

  Hide contents

 

This is so much fun and there's so much temptation of abusing it... I think there's a way to solve it, with the game trying first to push the new product above, but then not above again. At the top right instead, for example. Then at the center right, then bottom right, etc... Like a snail, Always getting larger. That could help to avoid this abuse.

[Thanatos994]

If they solved the teleportation bug in that way would it stop gases from reliably going up? Like, would it break filter-less SPOM builds? 

I think I'm confused because I'm not sure I fully understand what causes the bug--if the airflow tile in OP's design were a mesh tile, would the sour gas area flood with natural gas? Or would the natural gas go up anyways because it's less dense than the sour gas?

[Prince Mandor]

1 hour ago, Thanatos994 said:

f the airflow tile in OP's design were a mesh tile, would the sour gas area flood with natural gas? Or would the natural gas go up anyways because it's less dense than the sour gas?

Sour gas in airflow tile come in contact with cold tiles on left side. And sour gas became LIQUID Methane. Liquid cannot be in airflow tile, so it is pushed upwards. But there are solid tile, so it is pushed upward, and again and again and again. Until it meets any tile allowing liquid to exist in it.

This example demonstrates a lot of different tiles, pushing liquid up. Doors, windows etc.

[Kutuluk]

666 g/s natural gas output (666/90)*800 = 5920 W/s.

1020 W/s own consumption.

Net production 4900 W/s at 1 kg/s oil flow.

No external heat source is required.