Oxygen ventilation and pipe balancing

[Gus Smedstad]

I’m currently in the process of converting from a two-SPOM setup to a single 3-electrolyzer oxygen plant. I’ve been struggling for a while with distribution of oxygen, and I want something more efficient and systematic.

My dupes consume about 900kg of oxygen a cycle, and on a good day the two SPOMs generated 900-1000kg per cycle (1500-1650g/sec total). I want both to support more dupes and have more margin for error. My food supplies can support more, it’s oxygen that’s my current limit.

With the two SPOM setup, I had two ventilation networks, and it was fairly common for one vent system to fill up while the other was struggling to meet demand. I had a couple of crossovers where oxygen from one network could move to the other, but it felt like a messy and inadequate solution.

Given that I don’t seem to be able to load-balance the pipes, I want a system that will take the total oxygen output and distribute it to the pipes that need it. With an input of up to 2650 g/sec and 3+ output pipes, I don’t see any elegant way to solve that problem with basic pipes and gas bridges.

That’s really my question: is there a way to handle that much gas volume and reasonably distrubute it across multiple pipes, taking into account that pipes may fill up?

What I’m currently in the process of building seems like an inelegant solution: a room with 3 high-pressure vents and two gas pumps per outputput pipe. It’s kind of power hungry, 1.2 kw just for this central distribution point. I’m theorizing that the room will tend to run over pressure and gas will quickly move to the pumps that need it.

I’ve been doing OK for power, so I think I can afford that, though maybe I’m using my oil reserves too quickly. I’ve got several renewable sources I haven’t tapped yet - two oil reservoirs, a leaky oil fissure, and a natural gas vent. I’m using one natural gas vent now, the second one is just inconveniently distant from my base.

[Saturnus]

Why not use gas reservoirs? Have each SPOM go to a separate gas reservoir, and use the fact that if you make a T fork on the output of it the reservoir will alternate between the branches as long as there is room. Then splice each output branch from each reservoir into 3 lines.

[MorsDux]

Dont worry much about distribution. Exchange vents with high pressure vents plus atmo sensor plus shutoff. Set atmo to 4000 grams. you will not notice a 1000 g pressure difference anymore

[Gus Smedstad]

If I understand the pipe mechanics correctly, the output of a gas reservoir isn't special. The gas appears in the output square, and then follows regular pipe rules. Attaching two pipes to the output square is just like splitting a regular pipe.

Pipe splits are an annoying mechanic, since it always attempts each branch, whether the branch is full or not. If one is full the other branch gets every other gas packet, rather than the full flow rate of the pipe.

With enough gas bridges it's possible to set up a pipe split where gas packets alternate if both exits are clear, and all will go full speed down one branch if one is full, but it's a bit hairy. I've got one setup like that, and it doesn't quite work right - it drops every 4th packet instead of every 2nd. It's something to do with the timing.

Re: high pressure vents. Does that really work for distribution? My experience with using normal gas pressure is that oxygen doesn't move very far from the vent. I'd think that with a high pressure vent regulated to 4 kg instead of 2kg, you'd more movement and a larger area, but ultimately it would stall pretty much the way standard vents do.

I'm assuming oxygen distribution is a moderately well known problem, once your dupe count climbs moderately high and electrolysis is your oxygen source. Maybe not inevitable, since you can always keep your dupe count low enough so oxygen distribution never becomes a problem, but still something that arises often enough.

[Lilalaunekuh]

3 hours ago, Gus Smedstad said:

I want something more efficient and systematic.

Here is my scalable oxygen module:

Spoiler

I never had a false packet slip with this setup on start/load. NEVER (Can happen with some element sensor filtering setups.)

 

 

3 hours ago, Gus Smedstad said:

Given that I don’t seem to be able to load-balance the pipes

You could load-balance your pipes/electrolyzers but what´s the point of it ?

If your oxygen production exceeds your demand your able to keep all pipes filled.

=> You just need a network of gas vents to keep the pressure equal.

[Gus Smedstad]

I looked over your diagram, and near as I can tell you're not addressing the issue at all. Maybe I'm not following the piping, but it looks like there's no connection between the systems beyond combining the hydrogen output. Each electrolyzer feeds its oxygen output to a different pipe, and there's no provision for shifting gas from one to another.

As for the point, distribution affects production. In the simplest possible case, you could have one SPOM feeding a single air vent, and SPOM #2 feeding the rest of your vent network. Vent #1 will block quickly, and shut down SPOM #1. Now only SPOM #2 is operating, and it can't keep pace with oxygen consumption.

While my actual situation with 2 SPOMs wasn't quite that bad, I sure had cases where vent network #1 was backed up, and vent network #2 was starved, and my total oxygen production suffered.

Writing about this, though, made me realize that my whole "central pressure chamber" solution doesn't fix that at all. With my planned 3 networks, I could still run into the same problem, and shifting a little bit of air from network #1 to networks #2 and #3 at the start won't really help that. So I should definitely scrap that idea.

I need some sort of system where vent networks don't dead end. If there's oxygen remaining after it services all of its vents, that needs to feed into the next vent network. But recirculating it to the start won't help, since that's not where the shortage will be.

[Lilalaunekuh]

Yes, I´m not adressing the issue by merging my pipes.

(But I have more than one vent in a room leading to a separate "SPOM". If you go big enough the pipe itself will be your bottleneck.^^)

Like I said, what´s the point if your able to keep your all pipes filled all the time ?

(I never max out the potential output of an electrolyzer, so I am always able to keep up with load spikes.)

[Gus Smedstad]

If I'm understanding you correctly, your solution is massive over-capacity for oxygen production, and redundancy in your vent networks.

So, if my consumption is 1000 kg / cycle, instead of building 1100-1200 kg/cycle production, you build 2000 or 3000. The material cost of each electrolyzer module (one electrolyzer, two pumps, hydrogen filter) isn't that high. The cost is primarily space and labor to build them.

Over capacity doesn't cost extra water or power. Once your base is full of air and you're only replacing air the dupes breath plus filling new spaces, the water and power costs are going to be the same no matter how many extra oxygen modules you have.

You're also, if I understand correctly, building redundant vent networks. Some rooms - all rooms? - are serviced by vents from different networks, so if one is over taxed and starved, the second network will do the job.

I do know what you mean about pipe capacity being a significant bottleneck. Even two oxygen modules produce more total oxygen than you can carry with one pipe, and simple branches early in a vent network are bad because they cut pipe capacity in half. Better to use gas bridges so you fill a vent and the remaining 95% of the air keeps moving on.

[Lilalaunekuh]

4 minutes ago, Gus Smedstad said:

If I'm understanding you correctly, your solution is massive over-capacity for oxygen production, and redundancy in your vent networks.

Yes

 

5 minutes ago, Gus Smedstad said:

You're also, if I understand correctly, building redundant vent networks. Some rooms - all rooms? - are serviced by vents from different networks, so if one is over taxed and starved, the second network will do the job.

Yes, my "core base" got 24 vents, 3 for each of my electrolyzers. That´s the only space that needs oxygen vented inside.

Every other consumer needs oxygen inside a pipe (exosuit docks/oxylite refinery/...) or a fixed amount (self-build oxygen liquifier).

Spoiler

 

12 minutes ago, Gus Smedstad said:

I do know what you mean about pipe capacity being a significant bottleneck. Even two oxygen modules produce more total oxygen than you can carry with one pipe, and simple branches early in a vent network are bad because they cut pipe capacity in half. Better to use gas bridges so you fill a vent and the remaining 95% of the air keeps moving on.

But if you go big enough it´s about the breakpoint at which you need an other pipe.

Why should I bother and calculate with 1kg/s flow ? If I assume save 0,8kg/s and would need the same amount of pipes.

[Gamma17]

One thing that i tend to do is make all branch-offs (vents in case of oxygen) using bridges so that i get full flow speed no matter what and then feed it into the tank, which then feeds more vents. This is basically buffering and it allows to use 1kg/s of available throughput more effectively. Here is the example of 17 natgas generators (1530g/s) running on a single line (used it instead of oxygen since it is more compact and easier to screenshot):

As long as they are not used non-stop for too long (basically do not exceed 1kg/s average consumption over fairly large interval of time) they all run just fine. And it works great with oxygen distribution too.

[Nitroturtle]

7 hours ago, Gus Smedstad said:

Pipe splits are an annoying mechanic, since it always attempts each branch, whether the branch is full or not. If one is full the other branch gets every other gas packet, rather than the full flow rate of the pipe.

If you split off the output of a bridge, this won't happen.  When one line is full, the other will get full flow.  There's a bunch of other tricks with bridges you can use to prioritize flow that should help you achieve your goal.

In my bases, I usually use open air electrolyzers and only pump what I need for exosuits and circulation.  Makes for a much simplier setup imo.

[GemeinerJack]

i asked some time ago the same question and that was the solution we came up at that time: a 1-9 balancer

every of the 9 output lines had something around 11kg of oxygen from the 100kg input, so maybe you can adapt it to your needs

[Christophlette]

1 minute ago, GemeinerJack said:

i asked some time ago the same question and that was the solution we came up at that time: a 1-9 balancer

I remember this one. You had a big project because you needed 10 of those things 

[GemeinerJack]

Just now, Christophlette said:

I remember this one. You had a big project because you needed 10 of those things 

and the base got boring after i implemented 5 or 6 of them or so...

[Yunru]

A MOGOM has better conversion rates than a SPOM #ShamelessSelfPromotion

[ishakaru]

I don't have pictures, but what I did with my double electrolyzer spom setup(cycle 50ish) was to create 5 total output lines. 1:H2 4:O2. It will output hot O2 to 2 of the lines first then put the O2 through the weeze room.  It's a kinda a mess but it's 2 pumps to combine bridge, bridge to hot, overflow routed through weeze room. O2 chambers are separated while the H2 room is combined(2 H2 gens, 1 smart battery, 1 pump).

 

This setup is position center of the map several levels below my pod. For the cool O2 I have 2 vents per pipe positioned a distance from each other. Ideally bracing the living area. So the first vent releases into the central area and the 2nd one vents at the edges. From there I let entropy do it's thing while trying to keep CO2 levels down.