Sour gas help needed

[Jurij I. Gorkij]

Hello fellow dupes,

I ask for your assitance, please. In my current playtrough I went without teleporters. Because of that i want to take 10 kg/s Sour gas boiler to provide necessary clean water both for one dupe (oil driller) and up to 5 oil drills, so planetoid with oil wouldn't need any supplies from outside.

I looked for few designs (with 10 kg/s, so while i am aware of lower troughputs designs, I didn't want to go stacking them to 10kg) and all of them failed for me, while i wasn't able to solve it.

1) Tony advanced old design

I've built this in regular games, then started it... 20-30 cycles later, when temperature at aquatuners get to 750 Celsius (requested by temperature sensor), methan started to vaporize before it was pumped in correct chamber. Soon, sour gas chamber got overwhelmed by amount of natural gas.

Tried few tweaks, but were unable to get it working properly.

Next designs got built inside sandbox, but trying to start them regular way:

2) BrGustavoLS tweaked design

 

Tried this one too, even though it seems just a little changed version, same problem as 1).

3) EncSoup fool proof design (seems I don't fit in that category :-D)

 

I was unable to start this one - cooling chamber went too low temperature before aquatuner got hot enough. I know I was unable to set right settings to different sensors as their specification were missing from post. Hydrogen loop going trough cooling chamber got broken.

4) Leofarr design

Again, I wasn't able to start it up - cooling chamber ran out of any temperature while heating chamber got to only 200 Celsius.

----

To me it seems all of these designs were created in sandbox mode and started in sandbox mode, so authors weren't forced to deal with cooling/heating relative areas with limited resources (unavailable out of sandbox). But I am no pro, could (and probably will) there be specific details i am just missing.

What are designs you guys use for boiling 10kg/s of oil? And if you are using one from above, how is it working for you?

 

--------

 

Addition: own design

Trying to be constructive, I took EncSoup design as I liked idea of teleporting methan to it's own chamber. Redesigned cooling chamber so it would fit tepidizer (to eliminate overcooling problem). Provide additional way to get more cooling power by running coolant trough aquatuners in steam chamber (only if needed).

Start up went smooth, cooling chamber got right temperature quickly, crude went sour so more and more crude went in (controlled via hydrosensor). However after few dozens of cycles liquid vents were still open only for 60% of time, so I've added third aquatuner (as power is no concern, 10ks/s should provide about 60 kW). Somehow things didn't improve too much, so I went "screw this" and forced vents to remain open all the time.

Voilá, here's result (after left unattended for RL hour):

[Fradow]

Why not use a recent simple Sour Gas Boiler, namely @Yobbo 's one?

 

Note: I've not tested it myself, but Yobbo is someone I would trust to properly test his designs, and it's simple enough to figure out how it works from looking at it.

[Jurij I. Gorkij]

58 minutes ago, Fradow said:

Why not use a recent simple Sour Gas Boiler, namely @Yobbo 's one?

Note: I've not tested it myself, but Yobbo is someone I would trust to properly test his designs, and it's simple enough to figure out how it works from looking at it.

I've seen this design but i did disregard it for now since i would need to stack 3 of those to work with 10 kg/s.

[Yobbo]

I considered making a 10kg/s version, but it would have required another aquatuner in the boiler. In the end it was just simpler to prescribe as many of them as required for whatever the use case happens to be.

The ones that do 10kg/s with a single aquatuner are (as you found) quite fragile, reliant on tricks, and difficult to get to work outside of sandbox.

If you're making your own i'd recommend starting with that simple design, as everything else works on the same principles, just condensed and tweaked in various ways. That last Leofarr link you have is basically the same thing, just using more heat exchange media and a second aquatuner so as to handle 10kg/s. However as you found it lacks the safety features that would make it easy to start up and robust in case of failure. Probably just adding a tepidizer to it to heat the coolant during startup would at least get the boiler up and running.

[NurdRage]

I'm going to plug my own design here, and although i originally designed it in sandbox, i thoroughly tested various startup scenarios from survival. 

[Jurij I. Gorkij]

4 hours ago, Yobbo said:

I considered making a 10kg/s version, but it would have required another aquatuner in the boiler. In the end it was just simpler to prescribe as many of them as required for whatever the use case happens to be.

The ones that do 10kg/s with a single aquatuner are (as you found) quite fragile, reliant on tricks, and difficult to get to work outside of sandbox.

If you're making your own i'd recommend starting with that simple design, as everything else works on the same principles, just condensed and tweaked in various ways. That last Leofarr link you have is basically the same thing, just using more heat exchange media and a second aquatuner so as to handle 10kg/s. However as you found it lacks the safety features that would make it easy to start up and robust in case of failure. Probably just adding a tepidizer to it to heat the coolant during startup would at least get the boiler up and running.

Thank you, both for design (which is pretty nice) and for tip, however seems so far i have a winner

2 hours ago, NurdRage said:

I'm going to plug my own design here, and although i originally designed it in sandbox, i thoroughly tested various startup scenarios from survival. 

I've came around your design, however WARNING part made me look further for another one. This time i went and read it to the last paragraph. What really made me interest was amount of automation to prevent fail.

So I've tried to build it (just in sandbox), used regular resources for start up (as dupes would use - natural gas from previouis attempt, super coolant preheated to 85 Celsius). Currently running about 11 cycles (not counting heating of steam chamber), working like a charm so far. Thank you very much

I am planning to test it for few more cycles, then I think i have candidate for live game

[NurdRage]

1 hour ago, Jurij I. Gorkij said:

 

I've came around your design, however WARNING part made me look further for another one. This time i went and read it to the last paragraph. What really made me interest was amount of automation to prevent fail.

 

Awesome! glad it's working out for ya.

I originally intended the warning for players who didn't yet learn the game basics like automation, liquid flow, temperature limits, etc. I guess i should clarify :). As long as you've gotten to space on your own without the sandbox in survival, you should be okay.

I'll go back and adjust the warning

As for all the failure prevention, if it wasn't for that weird simulation glitch, i'd say the design is almost bullet proof. it runs smoothly for hundreds of cycles. If you encounter a rare failure mode, let me know so i can adjust. 

As For  those other designs you tried, i think a big issue is that their writeups don't include build-out-startup procedures. Nothing wrong with designing them in sandbox, but they must had start-up procedures to work in survival. In theory everything build in sandbox can be made in survival if the right startup procedure is implemented. Thus why the OSHA's revenge has a block-heater for startup. 

Maybe we should start asking that question with other builds, "Did you build this in sandbox? can this run as-is in survival? is there a startup procedure?" 

I think another issue is that as the game got patched, bug-fixed, and improved, little changes in the simulation mechanics gradually broke older designs. the Temp-swap bug altered the way many older boilers worked and broke them. So boilers need to be rechecked against new versions of ONI.

I imagine if they ever patch the liquid teleport bug, even my OSHA's Revenge will implode spectacularly. 

ah well. fun stuff. Thanks for checking out my design!

 

[Jurij I. Gorkij]

9 hours ago, NurdRage said:

Awesome! glad it's working out for ya.

I originally intended the warning for players who didn't yet learn the game basics like automation, liquid flow, temperature limits, etc. I guess i should clarify :). As long as you've gotten to space on your own without the sandbox in survival, you should be okay.

I'll go back and adjust the warning

As for all the failure prevention, if it wasn't for that weird simulation glitch, i'd say the design is almost bullet proof. it runs smoothly for hundreds of cycles. If you encounter a rare failure mode, let me know so i can adjust. 

As For  those other designs you tried, i think a big issue is that their writeups don't include build-out-startup procedures. Nothing wrong with designing them in sandbox, but they must had start-up procedures to work in survival. In theory everything build in sandbox can be made in survival if the right startup procedure is implemented. Thus why the OSHA's revenge has a block-heater for startup. 

Maybe we should start asking that question with other builds, "Did you build this in sandbox? can this run as-is in survival? is there a startup procedure?" 

I think another issue is that as the game got patched, bug-fixed, and improved, little changes in the simulation mechanics gradually broke older designs. the Temp-swap bug altered the way many older boilers worked and broke them. So boilers need to be rechecked against new versions of ONI.

I imagine if they ever patch the liquid teleport bug, even my OSHA's Revenge will implode spectacularly. 

ah well. fun stuff. Thanks for checking out my design!

 

Interesting - for testing purposes I left it to run over the night. I woke up to this:

As you said, vacuum/sour gas vent missing, so game paused. But interestingly, gas shutoff in the middle is broken because of heat (built from steel). Tried to reload oldest autosave (10 cycles ago), shutoff was broken already. Same goes with gas pump but i suppose that was just result of cooling not working around liquid vents together with increase of temperature.

[NurdRage]

Interesting!

What's odd is that only those things are broken. Usually if there is a breakdown in stability, a lot more stuff breaks. 

Could you send me the savefile of the unit so i can compare it against my own? 

[Jurij I. Gorkij]

58 minutes ago, NurdRage said:

Interesting!

What's odd is that only those things are broken. Usually if there is a breakdown in stability, a lot more stuff breaks. 

Could you send me the savefile of the unit so i can compare it against my own? 

It's really odd as i rebuild it in the morning and everything went smooth again - even though there was gas 600 Celsius coming trough shutoff, it was sufficiently cooled immediately. I've left it go for about 6 RL hours without any problems.

Here's save with what I came up to in the morning... One thing I am aware of is that i was cheapskate and used igneous rock as insulated tiles, however i wouldn't suspect that would be case of bug.

Boiler - Vacuum Bug.sav

[NurdRage]

hmm... i realize i didn't mention it in the post, but i made all the non-metal parts out of ceramic. The walls, insulated pipes, etc. I guess i'll edit that now.

Anyway, thanks for the save, when i get home i'll load it in and see what i can find.

20 minutes ago, Jurij I. Gorkij said:

I've left it go for about 6 RL hours without any problems.

This is what perplexes me. nothing can really accumulate over more than an hour only to break it later. If it's in steady state, it won't leave steady state on its own, something else had to come in and disrupt it. As you noted with the 600c survival, the unit can recover from pretty severe disruptions.

I wonder if you've found another really obscure simulation glitch even rarer than the one i found.

Maybe the drywall is holding more thermal mass than it should? maybe the igneous rock is bleeding heat that's accumulating somehwere? idk.

I'll check it out when i get home.

 

[Jurij I. Gorkij]

5 minutes ago, NurdRage said:

hmm... i realize i didn't mention it in the post, but i made all the non-metal parts out of ceramic. The walls, insulated pipes, etc. I guess i'll edit that now.

Anyway, thanks for the save, when i get home i'll load it in and see what i can find.

This is what perplexes me. nothing can really accumulate over more than an hour only to break it later. If it's in steady state, it won't leave steady state on its own, something else had to come in and disrupt it. As you noted with the 600c survival, the unit can recovery from pretty severe disruptions.

I wonder if you've found another really obscure simulation glitch even rarer than the one i found.

Maybe the drywall is holding more thermal mass than it should? maybe the igneous rock is bleeding heat that's accumulating somehwere? idk.

I'll check it out when i get home.

 

You didn't mentioned it, however I was able to realize that. My intention was to try it with igneous rock over night. Then switch to ceramic and let it be over day. Too see how big impact would that have. Since I woke up to damaged contraption, totally forgot about that.

Which is kind a happy coincidence, since boiler was going under same circumstances for another 120 cycles without obstruction so far.

I even have no doubt I will go with this boiler for live game  Next time, as I usually sit by the game while it plays, I could be witness of what happens as soon as overheating alarm goes off

[NurdRage]

I had a quick glance and found your problem, the very bottom thermo sensor, the one that measures the super coolant, is set to 240, it should be -240.

How it even works at 240 is actually quite amazing. I thought that was a critical sensor that would destroy everything, within 2 cycles, if set wrong. 

Apparently i built a lot more redundancy than i thought.

EDIT:

Wait a minute... something really weird about your particular save file is that the dev liquid sources, the thing that generates the oil, sometimes oscillates, like it generates oil every other tick, and then nothing. it might a bad interaction with my own mods, but it's really weird. 

check to make sure your oil flow is a continuous 10kg/s.

BTW, if you do fix the temperature thing, you may have to clear all the gas from boiler, it's at 1000kg/tile which is too much to handle. 

EDIT:

Another thing i noticed is that the bead pump boiler doesn't have enough interaction mass. My fault entirely, i didn't clarify in the post. But the central channel should have a steel radiant liquid pipe going down the middle that's not connected to anything. You can see it partially in the original pictures. Also a steel conveyor line goes down as well. Basically, cram as much interaction mass as you can. I put in the two things i mentioned and the outlet temperatures dropped by 30 celsius. 

[Jurij I. Gorkij]

1 hour ago, NurdRage said:

I had a quick glance and found your problem, the very bottom thermo sensor, the one that measures the super coolant, is set to 240, it should be -240.

How it even works at 240 is actually quite amazing. I thought that was a critical sensor that would destroy everything, within 2 cycles, if set wrong. 

Apparently i built a lot more redundancy than i thought.

That's funny... From my quick surveillance, coolant was staying around -230 - -220 without any intervention from automation. And as I've just had a look at your main boiler post, i was aware settings should be "above - 240 Celsius", it went wrong when i was setting the sensor (and didn't notice :-))

1 hour ago, NurdRage said:

Wait a minute... something really weird about your particular save file is that the dev liquid sources, the thing that generates the oil, sometimes oscillates, like it generates oil every other tick, and then nothing. it might a bad interaction with my own mods, but it's really weird. 

check to make sure your oil flow is a continuous 10kg/s.

I've noticed that too, it's reason why I have two of them there. Seems everytime, when flow in pipes get stopped for one tick (thanks to overpresure you mentioned), dev liquid source skips one tick of liquid once flow starts again.

1 hour ago, NurdRage said:

BTW, if you do fix the temperature thing, you may have to clear all the gas from boiler, it's at 1000kg/tile which is too much to handle. 

EDIT:

Another thing i noticed is that the bead pump boiler doesn't have enough interaction mass. My fault entirely, i didn't clarify in the post. But the central channel should have a steel radiant liquid pipe going down the middle that's not connected to anything. You can see it partially in the original pictures. Also a steel conveyor line goes down as well. Basically, cram as much interaction mass as you can. I put in the two things i mentioned and the outlet temperatures dropped by 30 celsius. 

I did add interaction mass (which isn't entirely your fault, as I said before, sour gas boilers are terra incognita for me). I've seen both white line between conveyor bridges and radiant pipe behind vents, but as inexperienced sour gas boilerist didn't know their importance.

So, I didn't clean pressure inside, thinking that break to cool the coolant will clear it (which it did). However that too started to overheat natural gas and gas pumps as flow was interrupted

What I really found hillarious and really good work from your side is the fact that even with those flaws and things i set up wrong, your boiler was still working at about 86% effectivity without major hiccups (judging from outputting gas pipes). What's more, I think that earlier failure I encountered was indeed that black hole tile, which wasn't detected because of wrong setup of my sensor (I suppose).

Thankfully, since I know how many my tests went wrong way, I have save from yesterday called "boiler start", where i am half way towards heating block heater. So I am currently preparing for another start before I'll attend to my dayjob again (letting game run in backround :-)).

[NurdRage]

Awesome! So that's it really, just fix the sensor to -240 and add the interaction mass (radiant pipe and conveyor line) and it should be great.

I never ran tests with igneous rock insulation. there might a tiny bit of performance loss since i tuned all the numbers for ceramic. Of particular importance is the oil heater temperature, 180 is for my setup. But i found changing things like condenser length or tile materials to push the optimum heater temperature a few degrees, sometimes its 192 for best results, sometimes its 178. But the performance changes were very tiny and almost impossible to notice unless you were measuring exactly the gas over at least 10 cycles of output (i had a whole measuring system set up to detect performance down to the 50g/s). chances are, you won't see any major differences.

14 minutes ago, Jurij I. Gorkij said:

 

What I really found hillarious and really good work from your side is the fact that even with those flaws and things i set up wrong, your boiler was still working at about 86% effectivity without major hiccups (judging from outputting gas pipes).

Thank you for your kind words!

I spent weeks tuning that design. I built it around a somewhat functional base and use automation to "nudge" it into an optimum operating zone. A "brute force approach" is less forgiving if the force applied is itself disrupted. I'm glad all those layers of safety and redundancy, both inherent and automated paid off!

hopefully it doesn't glitch while you're away. 

So safe it protects you from flaws in the universe!

[Jurij I. Gorkij]

45 minutes ago, NurdRage said:

Thank you for your kind words!

I spent weeks tuning that design. I built it around a somewhat functional base and use automation to "nudge" it into an optimum operating zone. A "brute force approach" is less forgiving if the force applied is itself disrupted. I'm glad all those layers of safety and redundancy, both inherent and automated paid off!

Kind words definitely well earned.

Interesting fact - when i first started boiler, some time after heating of steam void sensor went off, exactly as you said in your description. But then I had some issue (forgot to build something, don't quite remember), so I've reloaded and started again. This time void sensor was okay whole time.

I've just heated steam to desired temperature and I think void sensor went off again at exactly same time as my first attempt.

Sometimes I have feeling that sensors have tendency to set up different number when something specific happens (plus fact my mouse is kind of starting to break, interpreting click sometimes as double or holding the button as clicking etc.). So I kind have a feeling that sensor was set up wrongly only in one attempt, one we discussed and present in save.

And to not forget - thanks to @Fradow and @Yobbo for stopping by and trying to help me.

As my intent is (because of absence of teleporters) to build 10kg/s boiler on oil planetoid, providing enough clean water for electrolyzer for 1-4 dupes (one Loner is permament, then there's build team about 3 members which helps with building), for 3 pumps to supply boiler itself and use rest of water to pump additional oil and ship it to main planetoid. So 3 kg/s boiler would be possibly needed too

[NurdRage]

Would you like a 15kg/s boiler designed? and with the full-gamut of safety/stability features without regard to size?

[Jurij I. Gorkij]

12 hours ago, NurdRage said:

Would you like a 15kg/s boiler designed? and with the full-gamut of safety/stability features without regard to size?

Definitely interesting design to see...

[NurdRage]

 

This is a 20kg/s sour gas boiler i slapped together for ya. You can just delete whatever gas pumps you aren't using. It'll detect the overpressure and pause production until the pressure is relieved. The steam turbine is just there for heat deletion and no attempt was made for power efficiency. A lot of the safety systems are no longer needed since the steam turbine does most of the work of temperature control. 

 

Now because i built this in one day, i haven't fully battle tested it. No attempt was made to compact the size. And there is no oil heater or startup procedure. Another issue i've noticed is that it's a little too good in that the extra capacity makes it resistant to the vacuum tile glitch. The glitch still occurs, and it still reduces production, but it doesn't break the system, and the huge capacity means it still limps along producing some nat gas. Because production doesn't stop completely, the glitch detector doesn't work. So you'll probably have to implement specific detection methods for your colony. Like you detect the power output on your downstream natural gas generators is dropping. And that informs you that the glitch has occurred.

(EDIT: now that the glitch has been solved by another user, i'll revisit this design, the direct solution won't work since the condenser here is too big, but i think there are other solutions now)

 

 

The four aquatuners are thermium but the layout and design is very simple to follow and understand. If you want, you can only have one aquatuner be made of thermium and make the rest steel while moving them to their own separate steam box. it's more complicated but will save on precious thermium if you desire. 

 

Chonker 20kg v3.sav

 

EDIT: March 19th

Okay this new condenser design eliminates the glitch and properly collects sulfur. You're good to go. This will process the full 20kg/s of oil and can be adjusted downward by deleting the gas pumps you're not using. 

I've kept the design as simple as possible to understand and copy, so no optimization for size or stability has been done, 

 

 

 

Chonker 20kg v4.sav