Liquid Hydrogen Machine (No Debug Needed!)

[Quietghost]

I have discovered a way to make LH2 without cheats or the debug menu at all. Before we get to the actual machinery, there is some theory that we must work out.

 

First, the theory. This LH2-producing machine hinges on what could be considered a bug. When liquids or gases are fed through pipes at less than 1/10th their total capacity, state changes of the liquid or gas are ignored and the pipe remains undamaged.  Try it out! It's not hard to see where this is going. If we can build a machine to cool Hydrogen below its condensation point, but at the same time keep the pipe capacity below 1/10 of its total, then we get liquid hydrogen.

Now for the schematics for those of you who have a design already running through your head. We take room temperature hydrogen (oft referred to as H2), and run it into a primer that lowers the temperature of the hydrogen to ~ -246C. We then feed this cold gas into a condenser unit that utilizes the above bug feature to bring the hydrogen down below its condensation point of -252C. Ideally, we want to be well below that point to prevent any boil-off in the condensing chamber. We then pipe off the newly created liquid into vacuum storage so that your dupes can admire their creation!

Unfortunately, the actual machine is far from simple to build, requiring lots of pipes in awkward places and numerous valves to limit the flow and prevent pipe damage. We also don't want to bring the priming chamber all the way down to -246C. Instead, we may want to divert some of the gas from the cooling loop into the condenser when it is at the right temperature. This allows us to feed hydrogen continuously into the machine and run at steady state. The machine I have built follows the drawing below:

We first take in H2 at room temperature into the priming take. Using pumps, the H2 is passed through Thermo Regulators to bring the temperature down. It is important that the temperature of the hydrogen going into the TRs is not so cold that it is over-cooled in this first step. Ideally, we want the hydrogen exiting the regulators to be between -245 and -247C, so the input temperature will be T=(-245 + 14*#TR) where #TR is the number of Thermo Regulators you have on the priming circuit. For my implementation of 6 TRs, this works out to a temperature of -161C. This is where the heater is important. If the gas temperature is too low, then use the heater to warm it a little or input more room temperature H2 into the primer. Hydrogen that exits the TRs then passes through a Deciding valve (1). If the temperature is below -245C, then send it to the Primed H2 Tank. Otherwise, send it back to the priming tank to assist in the cooldown process. Primed H2 is then sent from its tank through a flow limiting valve, which limits the flow to 100g/s or less. A TR beyond that brings the temperature down to around -260C, and the gas is fed into the condenser chamber. Most of the gas will condense into liquid H2, which can be pumped out to a storage container or spilled on the ground, whichever you prefer. But some gas will not condense (see next paragraph) and will need to be removed from the condensing chamber with a gas pump. This gas is returned to the priming chamber.

As can be seen, the whole thing can get quite beefy if you want high production rates. There are 7 valves (1 for input, 6 check valves), 12 TRs, and a maze of pipes. It consumes 2.7kW of power and can produce 300g/s of LH2 when running steady. I built this with debug mode, but it can be adapted to be built by dupes instead. However, it is not as simple as flicking the switch and letting it run. The condenser has to first be cooled to LH2 temperatures, which requires feeding cold gas into the condenser and pumping it back out again until all the vents and pumps in the condenser are cold enough to not boil the LH2 off. In addition, you cannot simply set the check valve flow rate to 100g/s instantly. Because the TR takes a full second to boot, you have to start the flow at 50g/s and then ramp it up. Otherwise, the cold gas in the pipe will be above the 1/10th capacity limit and break the pipe. This means that those TRs must have a continuous power source or else the valves must be set permanently at 50g/s to prevent pipe damage. Also, once running and producing liquid, the LH2 removal pump must run continuously due to sporadic boil-offs of the liquid in the bottom. If the liquid is allowed to accumulate and enough boils off to push the pressure above 20kg/tile, then your vents will not work, your pipes with too-cold H2 will back up and break, and you will have a horrible, headache-inducing cleanup where dupes must get into the condenser and repair things, thus requiring another cool-down.

Hopefully this helps you on your journey to condense every gas! I have not yet found a practical use for LH2 in-game, but I'm sure someone will find one. Feel free to tune this design and find more efficient designs that don't create pipe mazes!

[Christophlette]

That's some nice setup !

I've only got a simple question : Did you try to go below 273.15°C ? (0K) I could try in debug mode but don't have the computer at hand at the moment and I'm curious ^^

[Saturnus]

18 minutes ago, Christophlette said:

That's some nice setup !

I've only got a simple question : Did you try to go below 273.15°C ? (0K) I could try in debug mode but don't have the computer at hand at the moment and I'm curious ^^

You can't. The game instantly crashes.

[Senfiron]

Have you tried to produce small amount of LH2 with this feature, and then use it in aquatuner for cooling remaining H2? Cooling 10 kg/s by 14C for 1.2 kWt looks far more effctive.

[Quietghost]

8 hours ago, Christophlette said:

That's some nice setup !

I've only got a simple question : Did you try to go below 273.15°C ? (0K) I could try in debug mode but don't have the computer at hand at the moment and I'm curious ^^

 

8 hours ago, Saturnus said:

You can't. The game instantly crashes.

Exactly, in fact it crashes for me even if I try 2K as a temperature. But it does not matter, since the thermo regulators only cool to a minimum of -268.9C no matter what. So you can make Solid H2 with this setup, but you can't liquify helium period

1 minute ago, Senfiron said:

Have you tried to produce small amount of LH2 with this feature, and then use it in aquatuner for cooling remaining H2? Cooling 10 kg/s by 14C for 1.2 kWt looks far more effctive.

That I have thought about, especially since the 1/10th capacity is now 1000g of liquid and hence much more cooling power. The problem is that the circuit needs to keep the flow rate at exactly 1000g in the critical temperature region (<-262C) or find some way to warm the liquid so that it may accumulate in a storage tank.

In theory, it can be done! But I have not yet figured out a safe, efficient system for doing so. Although I think I may have an idea...Ill try it tonight

[shanemadden]

16 hours ago, Quietghost said:

When liquids or gases are fed through pipes at less than 1/10th their total capacity, state changes of the liquid or gas are ignored and the pipe remains undamaged

Oh my.  Do they stay hyper-heated or cooled until released from the pipe (or collected into a block of >10% capacity)?  I feel like there's a lot of interesting potential use cases for this.

[Quietghost]

Just now, shanemadden said:

Oh my.  Do they stay hyper-heated or cooled until released from the pipe (or collected into a block of >10% capacity)?  I feel like there's a lot of interesting potential use cases for this.

That's right! So you could have LH2 at room temperature if you want, so long as it never accumulates above the 10% capacity. The moment it goes through a liquid vent, it vaporizes as normal.

[wachunga]

Here's some piping tricks you might find useful.

1) To get around the TR not starting immediately and requiring 50g packets at first, have the first packet run past the the TR and loop back into the feed stream. You can then set the valve to 100 g/s right from the start.

2) To handle that first 100 g packet looping back as well as giving you the ability to loop back packets in general (which can allow you to do other nifty things), use a shutoff to inject packets into that loop. Typically a bridge performs this function. A shutoff behaves slightly differently in that it gives priority to the loop (like a bridge) but it doesn't combine packets (unlike a bridge). So if there is already a packet in the shutoff's output pipe, the shutoff will not add anything to that packet. This maintains 100 g packets in that loop, only adding another packet when the pipe is totally empty.

3) To prevent the shutoff adding packets greater than 100 g, use a valve to feed the shutoff 100 g packets. To prevent packets combining in the shutoffs input pipe, have packets run past the shutoff and loop back into the valve's feed stream. Use a bridge to inject your supply gas into this second loop while giving priority to the loop.

This would be overly complicated to plug into your existing setup. Instead you could use these ideas with water and create a 1 kg/packet aquatuner cooling loop to condense the hydrogen. The aquatuner would move more heat and move it more efficiently. With a few other piping tricks, you can start and stop that aquatuner without combining packets into something that would break the pipes. See if you can figure it out.

[Soulwind]

While I do agree that it's pretty cool (see what I did there), I'm not really seeing much use for it.

You can't keep the liquid H2 cold enough to stay liquid without using the pipe exploit,

and to use the pipe exploit requires such small mass quantities that the net heat transfer

isn't much better than just using gaseous H2; pretty much eliminating the point of using a liquid

for the heat transfer in the first place.  Using it for drip cooling would be viable I think, but LOX

is so much simpler to create and maintain that I think it still wins for nearly all uses.

[The Flying Fox]

While I agree with @Soulwind and others that liquid H2 isn't terribly practical at the moment, it is a rather fun little challenge to build one of these things and actually getting it functional.  I decided to go a bit more conservative with my built to see if it could be condensed a bit and made easier to make in survival mode, if I so choose to do so at a later time.  It certainly wouldn't have the output of @Quietghost design, but it's certainly cheaper to run.

 

Instead of a priming room, I went with using various pipe/bridge/valve ninjutsu to mix various temperature packets together along with a small solid heat exchanger to add heat from the incoming H2 to slightly warm the H2 heading back to the primary thermo-regulator.  It doesn't quite run all the time, but nearly most of the time so there's a little room for improvement on the design.  One it was primed and working, I let it run for a few cycles and it seems to produce around 50g/s of liquid H2.  The starting temp of the hydrogen is 26.9C.  I did sorta follow along with @wachunga idea of using a bypass bridge on the secondary thermo-regulator so that it could process 100g packets.  Is obviously only saves power since it only seems to be doing around 50g/s.

 

 

The one thing I did note is the reason for 'failed gas' when it comes to producing LH2 which is shown in the second picture.  Even insulated abyssalite is not completely immune to the chilly effects of LH2   Although, it's mostly limited to that tile under the vent and the tile just to the left and in-between the LH2 and the petroleum.  It causes the LH2 to occasionally bubble out gas and the tile drops it's temperature fairly steeply.  From the looks of it, it happens at tiles where LH2 is forming.

 

I'm sure it's this effect as I know that abyssalite tiles/blocks are affected by it.  I'm just surprised that insulated abyssalite tiles are too. 

 

 

 

[Quietghost]

 

17 hours ago, The Flying Fox said:

The one thing I did note is the reason for 'failed gas' when it comes to producing LH2 which is shown in the second picture.  Even insulated abyssalite is not completely immune to the chilly effects of LH2   Although, it's mostly limited to that tile under the vent and the tile just to the left and in-between the LH2 and the petroleum.  It causes the LH2 to occasionally bubble out gas and the tile drops it's temperature fairly steeply.  From the looks of it, it happens at tiles where LH2 is forming.

Thank you for telling me about this tile issue, I confirmed that it was the problem. The tiles at the bottom of the condenser drop in temperature with each gas-off. So eventually they do reach LH2 temperatures and the gas-offs are no longer a problem.

[The Flying Fox]

14 minutes ago, Quietghost said:

 

Thank you for telling me about this tile issue, I confirmed that it was the problem. The tiles at the bottom of the condenser drop in temperature with each gas-off. So eventually they do reach LH2 temperatures and the gas-offs are no longer a problem.

Thinking about it just now, I guess the best way to solve it on a new build would be to use walls that are 2 tiles thick.  The inner layer being tiles with little thermal capacity that cool down to LH2 temperatures easily (Like metal tiles, probably gold be the best) and then an outer wall of insulated tiles either ceramic or abyssalite, or maybe just normal abyssalite tiles.  I know that regular igneous insulated tiles are effected by cold temperatures slowly, so that crosses out any of the normal materials.

 

On the other hand, I'd assume that liquid reservoirs sitting in a vacuum wouldn't have these issues as long as they're sitting on insulated abyssalite tiles.  Further thinking about the usefulness of this, I suppose if you want to be able to store a vast amount of hydrogen in a small place, this might be practical.  After all, the liquid reservoir can hold 33 times more in it then gas reservoirs.  Of course, my power strategy is usually just to spend my hydrogen and supplement my power with other generators.  Hydrogen, after all, is a pretty steady resource. 

[Kabrute]

if the energy in the tile could convert 5kg of material from one state to another it will transfer the energy regardless of insulative properties.  This is the cause of random ice melt drips in cold biomes and the cause of the above seen issue with trying to create super cooled liquids.  My lox builds are notorious for having to drop the temp of the containment unit first to the point that I eventually just lined it with metal then wrapped THAT with insulation..... in this way I end up precooling the metal while driving the system down to temp in the first place so that my liquids don't boil off on formation

[Glassyfo]

What colour is liquid h2?

[wachunga]

12 hours ago, bleeter6 said:

What colour is liquid h2?

Bubblegum.

I built a 1 kg/s version that takes the hydrogen down from 70C to liquid at -255C. Everything built in survival, only cheating I did was to give myself a bunch of 70C hydrogen to work with.

Spoiler

 

Spoiler

4 cooling stages to maximize energy efficiency. First stage uses an aquatuner with polluted water, hydrogen goes from 70C to -6C. Second stage uses an aquatuner with petroleum, hydrogen goes from -6C to -45C. Third stage uses thermo regulators, hydrogen goes from -45C to -241C. Fourth stage uses a 1kg/packet aquatuner with water, hydrogen goes from -241C to -255C and condenses.

Using a debug build, I calculated an overall energy requirement of ~5kW for 1 kg/s. First three stages are pretty simple. The last stage uses some trickery to ensure packets never combine and break the pipes. The last stage is also capable of stopping and restarting without problem.

Regarding the vaporization bug, I just closed the valve to the condensing chamber whenever there was a burp of vaporization. Once the hydrogen condensed again I opened the valve. Eventually the bug will stop once the insulated abyssalite walls get cold enough.

For funsies, I also calculated the cooling efficiency of each stage.

1st: 203C/kW. 31% duty cycle.

2nd: 86C/kW.  38% duty cycle.

3rd: 58C/kW. 100% duty cycle.

4th: 17C/kW. 66% duty cycle

[Quietghost]

In the new update branch I see that they are adding a hydrogen engine to the rockets. Maybe that can be a good use of the LH2 made with these systems, especially since that would be pretty close to reality.

I think in general that there should be some type of gas condenser building specifically designed for this purpose as well. It can liquify any gas, but only so long as the gas is within like 15C of the condensation point. Otherwise, to make LH2, we rely on this pipe exploit which, while pretty cool, is not what the developers intended.

[Saturnus]

1 minute ago, Quietghost said:

In the new update branch I see that they are adding a hydrogen engine to the rockets. Maybe that can be a good use of the LH2 made with these systems, especially since that would be pretty close to reality.

There's a super coolant in the update as well. No need for complex systems anymore. Just stick super coolant into an aquatuner loop and it'll cool to, but not below 1K. Almost too easy.