Electrolyzer producing wrong proportions

[Lifegrow]

3 hours ago, Nidhoggur said:

This build is kind of not optimal as well. Each of your electrolyzers should be overpressured half the time, since you seal them up with only one pump that pumps 500g/s, while the electrolyzer spouts twice as much

Basicaly, 3 out of 6 electrolyzers here are doing nothing.

I didn't say it was optimal  - I never aim for total 100% uptime of machines, for the most part I welcome them throttling themselves as @NanoD echoes below. Quite often I'd rather have a larger build that throttles itself in order to take advantage of passive cooling, especially with things like electrolyzers, thermo regulators, etc

1 hour ago, NanoD said:

500/s Actually is quit good, it is hard to get it to 600-700. you will basically never get to 800 if you dont put lots of pumps around.  And because it is idle about 50% of the time, it will only use 50% of the power. 

I think many gets in to the trap, it have to function on 100% or it is not good. But I think it is more interesting how much is the system drawing in power, how big it is and how complex is it to build, to how much is it producing.  

I seem to favor building big clumsy setups more often than not  

1 hour ago, Kasuha said:

Okay, I achieved perfect production rate 1:8 with this setup. It even works if the electrolyzer goes over pressure so removing the oxygen is not necessary, it can be left for dupes to breathe it out.

The upper half of the electrolyzer is covered by hydrogen. When a hydrogen bubble is released, it merges with that layer.

The lower half of the electrolyzer is covered by oxygen. When an oxygen bubble is released, it merges with that layer.

I like this, currently tinkering with something similar. I was doing an accept every dupe challenge, and was thinking about how I can possibly produce enough O2 without bottlenecking filters...This could be easily expanded on as well eh ?

[BlueLance]

3 minutes ago, Lifegrow said:

I like this, currently tinkering with something similar. I was doing an accept every dupe challenge, and was thinking about how I can possibly produce enough O2 without bottlenecking filters...This could be easily expanded on as well eh ?

I would assume it could be expanded, I think the limiting factor would be having only a one tile opening as there is more room for error? When my next colony dies I think ill fiddle with this one haha

[Nidhoggur]

2 hours ago, Lifegrow said:

I didn't say it was optimal  - I never aim for total 100% uptime of machines, for the most part I welcome them throttling themselves as @NanoD echoes below. Quite often I'd rather have a larger build that throttles itself in order to take advantage of passive cooling, especially with things like electrolyzers, thermo regulators, etc

I seem to favor building big clumsy setups more often than not  

Ok. I see, thanks for elaborating.

It does probably cool better, but it doesn't exactly save on power, as NanoD points out, since 6 electrolyzers@50% uptime is the same as 3 working constantly.

Aside from that, I sometimes end up building horrid convoluted systems as well.

[NanoD]

16 minutes ago, Nidhoggur said:

It does probably cool better, but it doesn't exactly save on power, as NanoD points out, since 6 electrolyzers@50% uptime is the same as 3 working constantly.

I am not just looking at the Electrolyzers I am looking at the hole system. To get a electrolyzer to run at 100% you will need to spend considerable  amount of energy on pumps.

[Kasuha]

31 minutes ago, NanoD said:

I am not just looking at the Electrolyzers I am looking at the hole system. To get a electrolyzer to run at 100% you will need to spend considerable  amount of energy on pumps.

I'm not sure if it is possible to run an electrolyzer at 100%, even if you pumps all around it.

[Nidhoggur]

6 hours ago, NanoD said:

I am not just looking at the Electrolyzers I am looking at the hole system. To get a electrolyzer to run at 100% you will need to spend considerable  amount of energy on pumps.

Um.. not really.
The amount of pumps would be the same as well. It's just that there would be 2 pumps for each EL instead of 1, but the number of EL's divided by half.
 

5 hours ago, Kasuha said:

I'm not sure if it is possible to run an electrolyzer at 100%, even if you pumps all around it.

I never tested it, tbh. But definitely more than 50%

 

[rezecib]

There was this post on Reddit about trying to get 100% uptime. They pretty much achieved it, although they could remove the hydrogen room to make it a bit more efficient.

[The Flying Fox]

9 hours ago, Kasuha said:

I'm not sure if it is possible to run an electrolyzer at 100%, even if you pumps all around it.

Well, thanks to the changes with the gas pumps in the last update, you can pretty much get them running almost at 100% with just two pumps.  There will be times when it drops down to 66% output for a moment or two, but otherwise its pretty stable.

 

Look how hot its running   And the oxygen output is about.. oh, 5 to 10 degrees less then that.  I tried to cool it with 4 wheezeworts on the sides, no good.  Didn't seem to drop the air temps much at all, so I put the wheezeworts in the base by the vents.  That actually seemed to work out okay.  Base has been sloowly warming up though over time...

[Sevio]

On 8-6-2017 at 1:27 PM, Kasuha said:

Okay, I achieved perfect production rate 1:8 with this setup. It even works if the electrolyzer goes over pressure so removing the oxygen is not necessary, it can be left for dupes to breathe it out.

The upper half of the electrolyzer is covered by hydrogen. When a hydrogen bubble is released, it merges with that layer.

The lower half of the electrolyzer is covered by oxygen. When an oxygen bubble is released, it merges with that layer.

I've been trying to get an electrolyzer room like this to work in a half open design, where the oxygen is free to flow out into the base. On top of being able to save energy on the filter, it would save even more energy on the pump for the oxygen.

Occasionally I've been having trouble getting them primed though, and sometimes oxygen makes its way into the hydrogen chamber afterwards. I understand that you're supposed to keep the oxygen pressure high enough to keep that from happening but are there other factors that can contribute to disturbing a hydrogen room?

I've built two of these units in my current savegame, made their rooms as small as possible so they would fit more easily:

They both destabilized recently and I just finished fixing the left one but it will probably destabilize again at some point.

So I experimented in debug mode with both a small and a larger version. In the first test I had the electrolyzer on the left side and both rooms destabilized at some point as I varied the oxygen pressure around them, and had a lot of trouble trying to get them stable again.

Then I built the versions you see below, with the electrolyzers on the right side.

I haven't had them destabilize yet over several tests, but I do wonder if that's just down to dumb luck or if it has something to do with hydrogen preferring to move to one side of a room.

I also did some testing with batteries on the hydrogen pumps and 100 kg of water to determine the pump efficiency. Making the hydrogen room bigger than the 2x2 minimal size in my first picture seems to help a bit with the pump efficiency, although the 2x4 room uses only about 102% of the minimum power needed to move hydrogen. The 4x4 room used slightly less than 100% of minimum power but that's probably down to a rounding error so I'll consider it 100%.

Would love to know if anyone knows of a more robust version of these, or if there is something I can do to condition the outside to make it more stable!

[Kasuha]

23 minutes ago, Sevio said:

Would love to know if anyone knows of a more robust version of these, or if there is something I can do to condition the outside to make it more stable!

I have this in my base. I spent some unnecessary time setting it up - since the pump chamber is so small, it actually cleans itself easily. I guess you can just build it, let the electrolyzer run for a short while with pump off, then switch it off to let the hydrogen settle, power up the pump so it draws remaining oxygen from the chamber, then power up the electrolyzer again. I had it destabilize once for a short moment after I added the wheezewort but it was perfectly stable ever since.

Note: it probably needs to be open at the bottom (below wheezewort) to keep the pressure stable and hydrogen from escaping from its layer. But the opening may not need to be as big.

[Sevio]

Looks just like the versions I was having trouble with, although it is a version with electrolyzer on the right just like the ones that worked well for me in testing. What do you have the atmo switch set to?

[Kasuha]

7 minutes ago, Sevio said:

Looks just like the versions I was having trouble with, although it is a version with electrolyzer on the right just like the ones that worked well for me in testing. What do you have the atmo switch set to?

Atmo switch is set to "above 1000".

[Sevio]

That's a lot higher than I would have expected, and I'm surprised it doesn't destabilize when the hydrogen pressure is that high! I've been trying to make do with settings around 500 or 400, which is why I needed bigger rooms to pump the hydrogen efficiently. I battery tested your version with the 2x2 room at 1000g for the atmo switch and it turned out to be the most efficient out of all the tests I've done, with only 98.8% of theoretical energy used for the hydrogen that was pumped. (for 100 kg of water, 12250 g of hydrogen was pumped, with an expected cost of 5880 J, but the battery lost only 5813 J)

I guess all I have to do in my survival base is flip one electrolyzer, increase the atmo switches and hopefully they won't destabilize again. Thanks for your helpful replies.

[The Flying Fox]

@Kasuha

 

Funny enough, the last day or so I was working on upgrading my base's oxygen and hydrogen production and came back to this thread and your design.  Here's what I came up with!

 

The nice thing I like about it (Besides not needing filters once they're properly purged of wrong gases) is that the concept is scalable.  Each unit can supply enough air for 4-5 Dupes and with the Atmo switches, can be self-regulating.  I had to lower the electrolyzers a tile because I was getting bubbles of oxygen getting into the hydrogen storage while getting them to stabilize.  (I got one to setup properly with the electrolyzer higher, but not the others, oddly.  Hence I lowered them, seems to still work the same.)

 

The two atmo switches that lead into the electrolyzers are both set at below 1200g.  One reads the hydrogen and the other oxygen and they basically shut off the electrolyzer if there's an over production of either gas and the pump lines are backed up.  The hydrogen's atmo switch is set to above 800g and the oxygen atmo switch is set to above 400g.

[Kasuha]

I replaced my old under-the-roof electrolyzer setup with two more and they seem to perform decently. For proper priming I needed to put the electrolyzers one tile lower and make the cuff one tile longer to let enough hydrogen collect before letting it to settle to single layer but after that happened, finishing the setup was easy. Remnants of the old setup can be seen all around, I'll wait till it cleans up remaining hydrogen under the roof, then I'll get probably rid of it.

 

[Sevio]

@The Flying Fox

Nice design, I can see how you might opt for pumping oxygen anyway if the base is complicated enough and you've got the power to spare. I might try for a second layer of hydrogen in the "cuff" area if I still end up have problems purging the hydrogen chambers. Since the electrolyzer shuts off automatically if it reaches max gas pressure and there is both a hydrogen and oxygen layer sitting on the electrolyzer, are the atmo switches to turn it off strictly needed?

@Kasuha

That's a good priming trick to know, my attempts at purging them always involved just messing with the atmo pump slider while the electrolyzer is running and hoping I got lucky.

[The Flying Fox]

4 minutes ago, Sevio said:

@The Flying Fox

Nice design, I can see how you might opt for pumping oxygen anyway if the base is complicated enough and you've got the power to spare. I might try for a second layer of hydrogen in the "cuff" area if I still end up have problems purging the hydrogen chambers. Since the electrolyzer shuts off automatically if it reaches max gas pressure and there is both a hydrogen and oxygen layer sitting on the electrolyzer, are the atmo switches to turn it off strictly needed?

@Kasuha

That's a good priming trick to know, my attempts at purging them always involved just messing with the atmo pump slider while the electrolyzer is running and hoping I got lucky.

I have loads of power to spare at this point.  Kinda one of things I like about this game.  You can have an abundance of nearly everything yet managing all of it is still a nice mental challenge.  And the pumps just make it easier to move the oxygen where it needs to go.  As for the atmo switches on the electrolyzer units, I actually thought that too.  Originally, I only had one switch on each unit that was right next to them. (I wanted a simple fast way to shut the unit off for whatever reason.)  But shortly after putting all these units in, I had -so- much hydrogen that it backed up in my feed line!  

 

Before I realized what was happening, the vacuum created by the oxygen pumps started sucking in packets of hydrogen.  Oops!    It wasn't many packets, actually, but it sorta defeats the purpose of making a seperator room like this in the first place.  Thus, I put the switches in to measure both levels of oxygen and hydrogen.  Also figured I might run into the opposite problem with the oxygen feed lines backing up and since the switches don't cost anything to run, why not?  Better to be sure that'll work properly.

[Sevio]

Thought I had them working but I came back to a broken hydrogen generator and a destabilized electrolyzer setup...

I notice there is 2 blobs of polluted oxygen hanging around the left (destabilized) electrolyzer room, could it be responsible for disturbing the hydrogen?

I'm a bit at a loss now for how I can stop this, there's something going on in this base that makes these things unreliable. Starting to think I might have to go towards a closed room setup and regulate pumping both oxygen and hydrogen even though that kills most of the efficiency gain.

[The Flying Fox]

1 hour ago, Sevio said:

Thought I had them working but I came back to a broken hydrogen generator and a destabilized electrolyzer setup...

I notice there is 2 blobs of polluted oxygen hanging around the left (destabilized) electrolyzer room, could it be responsible for disturbing the hydrogen?

I'm a bit at a loss now for how I can stop this, there's something going on in this base that makes these things unreliable. Starting to think I might have to go towards a closed room setup and regulate pumping both oxygen and hydrogen even though that kills most of the efficiency gain.

This does not surprise me.  From my experience with watching how the electrolyzer tends to output gas, it.. tends to be a little haphazard.  The unit seems to prefer outputting hydrogen from the left side of the machine, and oxygen on the right.  However, with tricks like we're doing here, it can be encouraged to unload the gases in particular tiles.  

 

Certainly any stray gases like polluted oxygen can disrupt this pattern, but what probably destabilizes them the most is just the ambient gas pressure around them.  If the pressure is too high near the unit, then it has to fight to push the new oxygen some where and that 'some where' might just cause a blob of oxygen to go between two blobs of hydrogen, then it migrates towards the hydrogen pump as it has no other place to go.  When the unit says 'Max Pressure', it's pretty much means there's no tile, at all, for it to output out a blob of gas.

 

Really, it all probably boils down to are the different amounts of hydrogen vs. oxygen that the machine creates.  Imagine that the two top tiles are filled with 500gs of hydrogen, but the bottom two tiles are nearly completely filled with oxygen.  (Close to 2Kgs)  In that case, it'll probably just shove aside one of those 500gram tiles of hydrogen into the other tile, creating a 1000g tile of hydrogen, and then dump the new 888g oxygen it just created into the now empty tile... which then gets pushed into the hydrogen pump because that 1000g hydrogen now wants to move towards the right.  

 

I would just try it with the electrolzyer lowered a tile with a two tile cuff, that might prevent random blobs of oxygen from trying to breakup the top layer of hydrogen.  If not, closed units.  Honestly, the 4 I setup are producing tons of hydrogen.  I had to install a second hydrogen generator although they can't run both gens 100% of the time, but that probably has more to do with the oxygen lines backing up then anything.  I need to print more Dupes!  

[Sevio]

I followed your advice, lowered the electrolyzer by one and have been doing some throughput testing with 1000 kg of water with both a closed and halfopen version. For counting the oxygen and hydrogen generated, I piped the resulting gases into abyssalite-insulated massive blocks of ice that instantly liquefies it.

The closed version (middle) I ran with your settings: pump hydrogen if > 800g, pump oxygen if > 400g. There's also an atmo switch to disable the electrolyzer if hydrogen backs up but I never ended up using it and I neglected to put in its corresponding switch for oxygen. Below are the stats:

1000 kg water processed (-0.5 kg left in electrolyzer)
115.5 kg hydrogen pumped
888.2 kg oxygen pumped
Total mass: 1003.7 kg pumped
Hydrogen Ratio: 0.130038 of expected 0.126126...

I also ran the open version (right side):

1000 kg water processed (+2 kg from previous run)
111.6 kg hydrogen of expected 112.2 kg hydrogen
Efficiency: 99.4%

I also tried to disrupt it by putting some high pressure oxygen tiles near it, as well as sprinkling polluted oxygen around and even inside the electrolyzer but it seems to be consistently stable. I think the result is close enough to consider it perfect, the only downside is that the 4 tile high version left just enough room to walk under it with my base's floor height and this one will block an entire floor. Time to rework my base!

[The Flying Fox]

@Sevio

Cool!  It's good to see that both units performed pretty much as expected and got just about the right ratios of gases.  I could certainly see using that open design myself as it's still fairly compact.  Good for sticking in smaller areas.  For either setup, you'll need a pipe to either take hydrogen away from the unit, or inject oxygen into the room from a remote closed unit.

 

Closed systems will probably have higher throughput per unit, meaning you need less of them and perhaps an easier power network setup.  However, I'm sure the air they produce is hotter then a comparable open system, and of course they use more power.  Heck, with those open units, I bet I could nearly run a base on hydrogen power alone!    Was nearly doing that during the thermal update and nothing has really changed.  Although, now, you don't really need a small army of fertilizer makers if you don't want too.  Sleet Wheat is quite fine without it.  

[Sevio]

I've been using the open units with lowered electrolyzers in my survival base for a couple hours and so far so good - they stay separated as they should and with only pure hydrogen needing to be pumped I'm making tons of surplus power with the hydrogen, far more than you would with a closed unit. Net surplus for closed is 296 J/kg of water and for open it's 722.24 J, 2.44 times as much. Or you can look at it as a 144% bonus.

Maths in the spoiler for the interested:

Spoiler

Making and pumping 1000 g of electrolyzer output in a closed system takes 120 + 480 = 600 J but only 112 g of that is hydrogen which gives you 896 J back. Net surplus is 296 J

Closed room with a single pump and a filter would take even more power, not sure how the filter power usage works but even if it runs only 50% of the time for the 500 g/s that a pump makes it's still a cost of 660 J for a net surplus of only 236 J. (80% of the power of closed without filter)

In an open system you only have to pump the hydrogen and the cost goes down to 120 + 53.76 = 173.76 J. Net surplus is 722.24 J, 244% of the power of the closed setup without filter.

 

[Grimgaw]

So I tried to combine efficiency with throughput (plus a bit of cooling) and came up with this:

Just 10 cycles of testing, but seems to be working. With pump being raised like this there's no problems with oxygen being stuck in there. 

[Sevio]

That's a very high oxygen concentration outside your unit! With the electrolyzer so enclosed and the air pressure outside being high, shouldn't the electrolyzer shut off? How are you getting its throughput that high?

[The Flying Fox]

Just now, Sevio said:

That's a very high oxygen concentration outside your unit! With the electrolyzer so enclosed and the air pressure outside being high, shouldn't the electrolyzer shut off? How are you getting its throughput that high?

Because the Wheezeworts basically act as pumps.  They suck in air from the bottom tile, and blow it out from the top tile.  They can pretty effectively channel gases in one direction, basically upwards.  I had a feeling they could be exploited for something like this 

 

Also, your math is slightly off, but the concept is correct.  The hydrogen pump doesn't spend 120 J/s for the 112gs, it'll spend it for whatever you set the atmo switch for, so it can pump full 500g packets into the pipe.  So, they're even cheaper to run.  The electrolzyer itself, many times, doesn't operate at full speed, but still uses 120 J/s in those cases.  In an open system, it probably runs at the 666 and 333 g/s more, then a closed system does, but yeah, an oxygen pump definitely soaks up a good chunk of power.

 

As for gas filters, they use 120 J/s for every packet (no matter the size of it) they process.  Packets move through at 1 packet a second through pipes.  So, if you send 500 gram packets through a filter, you're wasting half of its output for nothing in return.