deodorizers sucking polluted oxygen behind a liquid airlock

[anonimitazo]

I'm pretty sure this must be a bug but if it isn't, I've found by chance a very useful thing. What you see in the picture is a make shift room to purify polluted oxygen into oxygen, which I built inside of an ice biome in order to cool the gas + kill the germs faster. And to my astonishment, some of the ice condensed and created an airlock. But the deodorizer on the left keeps filtering the polluted oxygen when it shouldn't because there is water in between! which means that by doing this I'm able to take polluted oxygen from a room and turn it into oxygen in another room. But what is more surprising is the fact that the deodorizer on the right doesn't filter anything, and that is because it is not on the water. Should I report this?

[speckle21]

Already extremely well-known and wonderfully exploited to make UNLIMITED POWER  CLAY!!!!

 

 

 

[TheMule]

I've build this yesterday. 

tired to have random pufts drop slime around. There are a few g of CO2 that managed to creep in, but they don't prevent offgassing. More than a O2 production facility (surface too small to generate measurable amounts) it's just I want to force pO2 to be converted.

[TheOneFinn]

On 7/6/2020 at 6:25 PM, speckle21 said:

Already extremely well-known and wonderfully exploited to make UNLIMITED POWER  CLAY!!!!

 

 

 

Seeing as the original thread is locked and I cant reply there, a couple of methods are discussed about how best to toggle it on and off, any reason you cant just switch off the bypass pump? Automation wire to the pump vent (probably hooked to an atmo sensor on the o2 side) and it should stop pumping po2 causing the p02 above the pwater to overpressure and stop production with no resource build-ups. (Stopping the regolith/sand as suggested would not stop p02 production and build up?)

[Prince Mandor]

54 minutes ago, TheOneFinn said:

Seeing as the original thread is locked and I cant reply there, a couple of methods are discussed about how best to toggle it on and off, any reason you cant just switch off the bypass pump? Automation wire to the pump vent (probably hooked to an atmo sensor on the o2 side) and it should stop pumping po2 causing the p02 above the pwater to overpressure and stop production with no resource build-ups. (Stopping the regolith/sand as suggested would not stop p02 production and build up?)

You can switch off bypass pump, but it doesn't do anything with already pumped gas. If you have 1000 kg of po2 in airflow tiles they will convert to o2 after pump stopped. And this is a lot of o2

This contraption uses overpressurized pwater, so each offgasing will create tons of po2 and bypass pump take about half of in one pass. 

So my question is, what is purpose of turning it off? Possible best solution is stops getting away oxygen, not stopping.production

[nakomaru]

I think disabling the sweepers (or blocking access if done with labor) is a clean and good enough solution for most cases when you want to control the output pressure.

People want to control the output because you can remove the roof and supply to your base directly, but may end up with 150kg per tile of oxygen in your base before long.

[TheOneFinn]

1 hour ago, Prince Mandor said:

You can switch off bypass pump, but it doesn't do anything with already pumped gas. If you have 1000 kg of po2 in airflow tiles they will convert to o2 after pump stopped. And this is a lot of o2

This contraption uses overpressurized pwater, so each offgasing will create tons of po2 and bypass pump take about half of in one pass. 

So my question is, what is purpose of turning it off? Possible best solution is stops getting away oxygen, not stopping.production

First i'll preface my reply with i'm a programmer, coming from factorio, building complex systems is pretty much the sole reason I play any of these games. I have no fear of automation and am usually looking for an excuse to add more.

All my builds try to cope with any demand from zero to full in a graceful fashion, I dislike large buffers as all they do is hide resource bottlenecks (this definitely comes from factorio) and prefer all my production pipelines to operate in a "manufacture just as much as is needed" fashion with enough production capacity to meet peak demand. The only buffers should be to deal with intermittant supply (eg geysers)

As pointed out in the linked thread you need some form of throttle on it if you want to release the O2 directly into your base.

And put simply why would I want to turn pwater into o2 that I dont need?

48 minutes ago, nakomaru said:

I think disabling the sweepers (or blocking access if done with labor) is a clean and good enough solution for most cases when you want to control the output pressure.

People want to control the output because you can remove the roof and supply to your base directly, but may end up with 150kg per tile of oxygen in your base before long.

Disabling the sweepers will not stop p02 building up in the airflow tiles, or pwater continually being pumped in for no benefit, and removing it later will become a major pain in the neck with the po2 buildup, its going to be bad enough with the overpressurised pwater tile.

So I'm currently thinking remove one of the tiles next to the airflow to add an atmos sensor to the p02 area, once you hit say 2kg a tile of p02 you can stop both the bypass pump and the pwater input.

That way a) you dont keep unnecessarily adding pwater, if your production of po2 is able to meet demand you have enough water in the tile. b) your only risk of explosive decompression on removal is the pwater, which should contain the minimum level of pwater needed to meet demand.

So bypass pump and pwater only run when o2 above deoderisers is <1800g and po2 < 2kg per tile, you'll get a fair amount of wavering around that point (should still stay well under 3kg of o2)

[wachunga]

Conveniently, polluted water stops offgassing all by itself at 1.8kg gas pressure IIRC. So skip the bypass pump and you don't need to even think about polluted oxygen.

My take is that you'll eventually want to get the oxygen into pipes because natural gas diffusion is garbage. Plus normal vents overpressurize at 2kg so again no need to think about that, it just works. Once you are using pipes, you are free to seal the chamber and then you don't care about oxygen pressure there. If you care about overproducing oxygen, then just use a valve to limit the polluted water coming in to match your dupe count. Dupes holding their breath reduces total oxygen consumption, because ONI physics, so you can't balance it perfectly but who cares. 

The funny thing is that as broken as this seems, electrolyzers are more broken because they make oxygen and power!

[TheOneFinn]

5 minutes ago, wachunga said:

Conveniently, polluted water stops offgassing all by itself at 1.8kg gas pressure IIRC. So no need to even think about polluted oxygen.

Only off of the pwater, thats being pumped into the airflow tiles by the bypass pump so as long as that bypass is running the pwater will keep offgassing and adding more po2 to the reservoir of po2 inside the airflow tiles.

The purpose I was trying to acheive was to gracefully throttle production so that it matches demand, even if demand is signficantly below the maximum production with no unnecessary buildup or resource usage. So far the replies have all been along the lines of "why would you want to do that?" rather than potential problems in the solutions I proposed. 

Personally I find the resistance to even think about throttling production to match demand, baffling. You never just keep producing steel in factorio with nothing to use it.

 

[wachunga]

I worded that poorly and made a bunch of edits after the fact. My thought was to not use the bypass pump at all, just have overpressured polluted water if you like overpressure tricks. Or a wide polluted water basin if you don't like overpressure tricks.

[TheOneFinn]

25 minutes ago, wachunga said:

I worded that poorly and made a bunch of edits after the fact. My thought was to not use the bypass pump at all, just have overpressured polluted water if you like overpressure tricks. Or a wide polluted water basin if you don't like overpressure tricks.

That will signifcantly cut down on production though, since po2 offgassing is dependent on gas pressure you actively want to keep the area above the pwater at very low pressures.

I dont understand though, why is my idea of an controlling production via a pair of atmos sensors bad? It doesnt even need any more tiles since you could simply move one of the tiles below the airflow sensors down to add a one tile gap to get a pressure reading?

 

42 minutes ago, wachunga said:

then just use a valve to limit the polluted water coming in to match your dupe count. Dupes holding their breath reduces total oxygen consumption, because ONI physics, so you can't balance it perfectly but who cares.

Thats what I dont like, "fixed" builds, I much prefer something that actively monitors current demand and alters production dynamically to compensate.

[wachunga]

I don't think anyone is saying controlling production is bad. Rather it's an extra step that most people don't care about. The resources involved are so far from scarce as to be practically infinite. You enjoy playing the game in a way that closely matches production to demand, most people only care when something is scarce.

[TheOneFinn]

5 minutes ago, wachunga said:

I don't think anyone is saying controlling production is bad. Rather it's an extra step that most people don't care about. The resources involved are so far from scarce as to be practically infinite. You enjoy playing the game in a way that closely matches production to demand, most people only care when something is scarce.

Sure I get that, but its incredibly frustrating when it seems like people refuse to even acknowledge there is a problem that could be desirable to fix, let alone discuss the merits of different solutions. I'm not trying to convince anyone that they should switch to a more economical playstyle if they dont want to, but equally why do I have to fight to justify an efficent logistic and production system?

As a programmer, its the different between robust, encapsulated and reusable modular code, and a quick hack that solves the problem but will be a maintenence nightmare along the line.

Why would you want a system that I had to manually adjust when I took on a new duplicant when a "fire and forget" build that will automatically adapt (and even play an alarm if supply is unable to meet demand) is achievable? I guarentee I'm going to forget to do so, especially with 5 other minor crisis going on. 

I'm not trying to be provocative or argumentative, I just want to discuss solutions that actually solve all the problems is a robust manner rather than half-way measures that solve only some of them or having to argue why these problems need solutions in the first place.

[nakomaru]

@TheOneFinn Sorry, I wasn't intending to deny your playstyle. I merely did not recognize large PW mass as a problem to be solved. I thought it was just about reliably controlling the pressure.

If you toggle both the bypass&PW only when PO2 & O2 are below target, you are still left with the the problem of large PW mass if you are consuming less than 9kg/s. You could calculate the necessary PW flow, but that's not dynamic.

Solution 1

Solution 2:

1. Atmo sensor in PO2 area, which adds water when below 20kg/tile.
2. Unthrottled bypass.
3. Regolith regulated O2 production.

[TheOneFinn]

1 hour ago, nakomaru said:

 

1. Atmo sensor in PO2 area, which adds water when below 20kg/tile.

Thats basically what I proposed above, but does it need 20kg/tile of po2? From what I could gather deoderizers are not pressure dependent?

If the deoderizers are unlimited they will keep converting po2 even if we stop the bypass pump to stop its production, so even if the we stop the bypass pump when o2 pressure reaches the desired threshold it will still continue to overproduce by whatever po2 is buffered in the airflow tiles.

Thats why I was thinking of limiting the po2 with its own pressure sensor to less than 2kg so you have at most 2kg x width o2 overproduction when you hit the o2 pressure limit.

Secondly by stopping the pwater being added whenever either o2 or po2 are at a "reasonable pressure" presumably we'd only every have just enough pwater to match the po2 demand (which is dependent on o2 demand). Probably having the bypass pump hooked directly to the sensor with maybe something like a 30s filter on adding pwater, either that or 2 different pressure limits for bypass and pwater

We dont actually care about how much pwater is in that tile, or the flow nescessary, only that we are producing at least as much po2 as nescessary which can be ascertained by po2 pressure? (What we really want is the rate of change of o2 pressure, but Oni's automation system doesnt allow that )

The only issue i see is that the "ramp up time" is limited by how fast you can add pwater, a sudden massive spike in demand is going to require time for pwater to be added to ramp up production, but so long as you do that fairly regularly it should stay fairly close to the maximum limit (assuming you are nowhere near needing 10kg of pwater per second). The natural latency in the system is going to tend to mean you will have more pwater than required and a less than 100% uptime on the bypass pump I'd have thought?

 

[nakomaru]

1 hour ago, TheOneFinn said:

We dont actually care about how much pwater is in that tile, or the flow nescessary, only that we are producing at least as much po2 as nescessary which can be ascertained by po2 pressure

?

Spoiler

 

3 hours ago, TheOneFinn said:

The purpose I was trying to acheive was to gracefully throttle production so that it matches demand, even if demand is signficantly below the maximum production with no unnecessary buildup or resource usage.

3 hours ago, TheOneFinn said:

Disabling the sweepers will not stop p02 building up in the airflow tiles, or pwater continually being pumped in for no benefit...

...

That way a) you dont keep unnecessarily adding pwater, if your production of po2 is able to meet demand you have enough water in the tile. b) your only risk of explosive decompression on removal is the pwater, which should contain the minimum level of pwater needed to meet demand.

Anyway, I can see how your solution will also eventually reach a steady state for PW mass, because eventually one off-gassing event will be more water than you will take in waiting for an off-gassing event. I would guess my solution would find a lower minimum PW mass because it will always be able to off-gas.

You may want to check for a higher PO2 mass (like 20kg) in order to prevent waiting for an off-gassing event.

[TheOneFinn]

36 minutes ago, nakomaru said:

?

  Hide contents

 

I mean the actual "raw amount in kg" of pwater is irrelevent, and is purely a function of desired PO2 production, which is itself a function of desired O2 production, the quantity we care about is final O2 production, the rest are just "whatever value is needed to meet that" so not numbers that we actually care about. 

I'm also not sure you have understood what I meant from my reply

I propose moving one of the tiles under the po2 storage air tiles down to add in a one tile gap to put an atmos sensor in, this would disable the bypass pump vent when po2 is above say 2kg and disable pwater being added unless po2 is below the level for say 30 seconds.

Assumption: the bypass pump can VERY quickly move po2 from just above the pwater into the airflow tile reservoir. Thus pwater offgassing can be fairly accurately controlled and that offgassing is reasonably closely tied to the bypass bump operation.

It might overshoot on pwater initially, but then it will shut down the input and off-gassing will naturally decrease the amount, initially it will overproduce slightly causing the bead pump to run intermittently, but as the pwater mass decreases the bypass pump is going to run more and more often until the pwater is "maximally off-gassing", the po2 still isnt reaching the threshold so we start adding more pwater, thus it will always vary between very slightly below and some value above the required amount, but over time the average should be within a reasonable margin of error of the minimum required.

[Prince Mandor]

12 minutes ago, TheOneFinn said:

I propose moving one of the tiles under the po2 storage air tiles down to add in a one tile gap to put an atmos sensor in, this would disable the bypass pump vent when po2 is above say 2kg and disable pwater being added unless po2 is below the level for say 30 seconds.

Assumption: the bypass pump can VERY quickly move po2 from just above the pwater into the airflow tile reservoir. Thus pwater offgassing can be fairly accurately controlled and that offgassing is reasonably closely tied to the bypass bump operation.

It might overshoot on pwater initially, but then it will shut down the input and off-gassing will naturally decrease the amount, initially it will overproduce slightly causing the bead pump to run intermittently, but as the pwater mass decreases the bypass pump is going to run more and more often until the pwater is "maximally off-gassing", the po2 still isnt reaching the threshold so we start adding more pwater, thus it will always vary between very slightly below and some value above the required amount, but over time the average should be within a reasonable margin of error of the minimum required.

 

In theory you describing working system. I don't see errors in your description.

In practice, as soon as you start using hydrogen rockets, oxygen became a waste byproduct, not needed any more. Also, as you can see, this design called Claymator. It's main purpose to produce clay, oxygen is byproduct. On technologically advanced base your need some way to destroy oxygen or to store infinite amount of it. Of course, you can use same automatics based on clay storage, not oxygen pressure.

 

Another point. Offgasing is random process. You can have a streak of bad luck. I, personally, had once situation of no offgasing for cycles. So, this randomness needs some buffer to compensate. If I remember correctly, it is 0.1% chance per tick and converts 0.1% of water mass. Lets think about numbers. 5 ticks per second, it is averaged one offgasing per 200 seconds and you need 111.11 g/s poxygen per duplicant. It means  22.22 kg per offgasing and 22 tons of water per duplicant.

Let's think about 10 duplicants. It means 220 tons of pwater packed in a cell randomly creating 220 kg of po2. bypass pump takes 110 kg away in first moment, so this system jumps from 0 kg to 110 kg in second.

Is it really possible to stabilize it by measuring po2 pressure? I don't think so

With chances like 1/1000 you nearly never have perfect situation of exactly 2 kg of po2 in airflow tiles zone. it can be a minutes of total vacuum or thousand kilograms if chances rolled quickly one after another. This is unpredictable and extremely volatile.

This system throttled by deodorizers. each deodorizer provides 90g/s of oxygen. so you need 11 deodorizers per 10 dups. Is it possible to control it by deodorizers? I don't think so. If you have filled deodorizers, it means you have 320 kg of sand in each, and this amount enough for 2400 seconds of work after you turn sand delivery off. Possibe, but 216 kg of oxygen per deodorizer will be overproduced.

If you stop water delivery, you stop delivery of 10 kg/s to a buffer of 220 tons. You can stop it in any time simultaneously with any other way to stop Claymator -- it is not precise by any means.

So only way we have is close oxygen area, and measure oxygen going to base from this system, starting/stopping waterintake simultaneously with air pumps. This entire system is enormous buffer in each of its part, so we can only limits it by controlling output. All other ways don't looks meaningful.

 

But of course you can build it any way you like, and looks like it will be. Possibly you will create a great design, or found some useful tricks in process. Practice is a best measure

[TheOneFinn]

1 hour ago, Prince Mandor said:

Another point. Offgasing is random process.

Thankyou, this is the detail I was missing, I hadn't quite realised the consequences of all of the pwater being merged to one tile.

That would presumably be mitigated with the a larger number of sources, the pwater source is 6 tiles wide, so you could potentially fit 2 in for your 10 (12) deodorizers for your hypothetical dupes but by the sounds of it 2 would not be enough to guarantee continuous production. We could expand the pwater tile side ways to the full width of the build, as I understand it average output is based on mass so it doesn't matter if its in one tile or several, the more tiles the closer to a continuous stream of average production we would get.

I suspect your right and that O2 output is going to need to be limited too.

I'll admit I am more interested in this as an early to mid-game build than end game, something that can be built for 30 dupes when you have 6 and then forgotten about, that uses the minimum of pwater and produces just enough oxygen whilst producing a steady supply of clay for ceramic. That needs almost zero power initially but can be expanded by adding gas pumps over the deodorisers to feed to suit docks once researched and built.

 

[nakomaru]

And, when you are waiting for an off gassing event, you are pumping in 10g/s of water the whole time.

This is why your solution will tend towards to find a maximum amount of PW mass. One off gassing event will need to give off e.g. 200 seconds * 10kg/s = 2T of water while it was waiting (and also not producing) to reach a steady state. That's 2000T in the PW tile.

Solution 2 however, keeps vacuum for your tile to always be able to off gas, and only adds water when the mass is small enough that it can't keep the PO2 room full. This should find a much smaller PW mass and only needs one tile. It will overproduce a few tons of oxygen each time you turn it off due to stored regolith/sand, but that may not be a problem depending on what you set your sensors to and how big your base is.

[TheOneFinn]

1 hour ago, nakomaru said:

Solution 2 however, keeps vacuum for your tile to always be able to off gas, and only adds water when the mass is small enough that it can't keep the PO2 room full. This should find a much smaller PW mass and only needs one tile. It will overproduce a few tons of oxygen each time you turn it off due to stored regolith/sand, but that may not be a problem depending on what you set your sensors to and how big your base is.

I think at the very least I'd put the pump on a buffer timer, so it switched off after a significant amount of time of no demand. Otherwise the default state is to convert all pwater as fast as it can.

Chances are when it comes to decommissioning, at least when I'm playing, there will be a significant delay between the build of its replacement and its tear down, left long enough all the pwater will be converted and become significantly more of a pain to relocate as a gas than a liquid. Given that its literally just a vent to be automated it doesnt seem like it would hurt to support long-term suspension.

I'm not sold on filtration control as a means of regulating o2 production, it just doesnt seem like it would be capable of maintaining any kind of accurate pressure. I think you can put deodorisers on doors for instant control (I know it works for terrariums and seemed to control pressures to within 100g or so) but you lose the neat po2/o2 seperation. Meh!

[beowulf2010]

I think you're now seeing why most of us don't bother trying to control it. 

[nakomaru]

7 hours ago, TheOneFinn said:

left long enough all the pwater will be converted and become significantly more of a pain to relocate as a gas than a liquid.

It will not off gas forever. It will off gas to about 1T per tile.

Either you think 20T of gas is harder to deal with, which could be dealt with in several low effort ways (such as waiting to turn it off for a few cycles), or you think 2000T of water in a single tile is harder to deal with than 100T (these are my ballpark guesses as to the steady state mass of of the water tiles in the designs - it would be a lot lower than 2000T depending on how appealing you find multiple tiles).

If you have a very small base, controlling via filtration medium is not going to work. But if you have at least 1000 tiles of air, you ought to be able to control between e.g. 1.8kg-2.8kg with 10 deodorizers. (The average remaining medium should be about half of the maximum, so 160kg of medium → 100kg of oxygen per deodorizer).