Liquid shutoff does not close despite red automation signal, even without power

Hi @zirrboy!

I don't believe anything is wrong here. There are some things to take into account going further.

A water sieve can process up to 5 kg/s of polluted water, so half a pipe's worth at any given time. If there is any kind of interruption in, say, providing filtration medium to the sieve so it can operate normally then its input pipe will stop receiving liquids and stall the flow into it. This may be creating the backing up that is seen in the first screenshot.

The second screenshot shows that the shutoff's output pipe is blocked and what I described above is mostly what's happened. Once liquid finds itself on the output side of the shutoff it's no longer in its control. The visual effect that's seen is that liquid continues to flow, but that will only be for the remaining amount of liquid in the pipe on said port.

Then again, I am working off static images here. It'd be great if you could provide a video showing more of what's going on.

Food for thought:
Water sieves can be automated to throttle production instead of using the liquid shutoff, this will avoid this particular situation you're describing.

Yeah, a video would've probably been better, @JRup. The screenshots were taken at different times and it didn't occur to me how confusing that would be. From what I understood from your comment you suspected it was simply backed up fluid already past the shutoff that kept flowing. And my screenshot does indeed only show that side moving. But the issue is that the shutoff keeps sending packets of fluid through.

I am seeing the exact same bug: liquid shutoff with automation signal set to red still allowing partial flow (my consumers after the shutoff consume < the 10kg/s liquid pipe maximum). in 1 case the consumers are oil well.s in another the consumers are thimble reeds. i can also temporarily work around the bug by deconstructing and reconstructing the shutoff, but eventually the bug manifests again.

I am experiencing the same issue. 

 

In my case I have 2 shut offs in different parts of my starting asteroid that sometimes do not shut off after loading the save. They both feed petrol boilers that will catastrophically fail when the shut off doesn't shut off. One will fill completely with crude, and the other will overflow and flood the base with sour gas, if they don't shut off. So, I have made it a habit to watch them after each load. The fix is simply to deconstruct them and reconstruct them and then they work fine until the next load when they might fail again.

 

In each case the arrangement is a simple pipe of crude into the shut off, then a 2-tile pipe to a valve after the shut off to restrict the rate of flow into the boiler. There are no parallel pipes or loops involved. The shut off is controlled by the signal from a hydro sensor that is green when there is no liquid 1 tile away from where the crude drips. If that hydro sensor goes red and the shut off doesn't shut off, the boiler floods and overflows.

 

I know the pipes are correct, and not joined behind the shut off valves because I deconstruct and reconstruct the bugged-out shut offs on a regular basis. 

 

I didn't notice this happening before the recent update, but then I was building this base when the update dropped, and the boilers might not have been able to overflow yet.

 

Edit: I don't know if this makes a difference, but I play on both Linux and Windows, and use the Steam cloud to load the save from each system. This save has been back and forth between Linux and Windows many times. 

Edited April 17 by B.Y.
Additional information

Came here to also report the same. It seems to only happen if the output pipe has a partial packet, as seen in the attached screenshot - it's reporting "pipe blocked" yet there is still a small amount flowing.

Deconstructing the liquid shutoff and immediately rebuilding it stops the flow, as others have said, so it's not a plumbing issue.

I've since reworked this part to use a liquid meter valve instead, which is more suitable for my scenario anyway, but that needs plastic, and obviously, the regular shutoffs passing partial packets is still a problem.

Same thing here. Have seen this happen multiple times in recent weeks.

Consequences might range from benign to disastrous next time this bug affects one of my shutoffs, depending on which it is.

Seeing the same thing here, the key liquid shutoff valve that was making sure my toilet water got filtered through chlorine before entering my water supply now is broken open and letting all liquid through. It had been broken for a while as it was slowly draining my three full containers of polluted water out. Looking back through autosaves it broke sometime in my play time yesterday, though I don't have the save from before it broke. It may have broken when I went back several different autosave files to undo a horrible disaster in something I was building.

I hadn't installed any new mods, and the only mods I have are small visual ones that increase precision of the temperature display of fluids. 

I'm on Mac, so this doesn't seem to be system dependent.

Edited Wednesday at 10:40 AM by ergzay

!!! Oxygen Not Included: March 2026 Bug Fix Update

https://store.steampowered.com/news/app/457140/view/535505450660003889

The Liquid Shutoff will now retain the current flow it was saved with, rather than resetting to 0 on load.

 

Hello! I have the same problem with the liquid shutoff valves, as well as the gas one, and I just noticed that I'm starting to 'leak' the hydrogen storage, which is absolutely unacceptable because it creates a 'game-breaking bug'.

 

The last time I played for any length of time was around October-November 2025, and everything was fine then. On my second asteroid, I had a water aquatuner 'drip cooling' system set up, using a 500mg dispenser and an electronic valve that was disabled based on certain settings. Everything worked fine for months.

About a month ago, I returned (originally savegame to the Space Out) and quite by chance noticed that the shutoff continued to let in water at the rate allowed by the dispenser (500mg), even when the shutoff was electronically disabled. I noticed this when the reserve water tank was empty, and the aquatuner overheated. This is abnormal.

If I disassemble the shutoff and build a new one, the problem goes away, but if I save and load it several times, all the problems return (this is very easy to reproduce).

It doesn't matter what the configuration is; I've tried moving the dispenser, installing two dispensers, etc.

 

I recently assembled a system with a hydrogen generator and a hydrogen tank, which is activated via an electronic gas interlock. I have the same problem: the interlock is disabled, but the hydrogen flowing through it is exactly the same volume as the hydrogen generator burns.

Simply saving and loading the program doesn't fix the problem.

 

That is, it's easy to reproduce not only on water shutoff valves but also on gas ones. The first three lines of my message and the probable (in my opinion) cause of the incident. Let me remind you that this problem certainly didn't exist in October-November 2025.

 

I ask you to resolve this issue as quickly as possible, as it creates situations where it becomes virtually impossible to play (I can't manually disable and enable all the tanks using the scissors, and without a blocker, the resource consumption is absolutely unacceptable).

Oddly, I didn't see this complaint on the Steam forums.

 

I didn't use mods, but when this issue first surfaced (after reverting), I hadn't installed any mods at all.

 

I think the problem arose either because of the upgrade to the new Unity or because you started storing the last values on the gas-water shutoff, because when they were reset to 0 on 'OnLoad,' this problem clearly didn't occur.

 

The situation is the same as in the second screenshot (bughouse26 Posted April 17)

https://forums.kleientertainment.com/klei-bug-tracker/oni/liquid-shutoff-does-not-close-despite-red-automation-signal-even-without-power-r52156/?do=findComment&comment=73287

I just reproduced the problem (I hope it gets fixed quickly).

 

I have an aqutuner in the vacuum of the upper part of the second planet, Space Out. It's in three water cells. As soon as the aqutuner is turned on, the electronic water valve activates, and water starts dripping 200 mg at a time through a dispenser (to prevent the 10 liters in the cell from dripping) into the cooling reservoir (I have a lot of water).

 

1) I disassembled the Liquid Shutoff and reassembled it, turned it on and off. No problem. The lock works as normal.

2) I saved one game, loaded it, no problem. I saved (2 save), loaded it, no problem. I waited a day, saved it, no problem. I exited the game and loaded the last save, no problem.

3) When the Aquatuner turned on, the Liquid Shutoff (with the 200mg dispenser) turned on, and I saved!

I loaded it, and water was leaking through the Liquid Shutoff, even though it's turned off.

 

This means that Liquid Shutoff remembers which portion it sends per tick (if a save occurred at that moment), but completely ignores the fact that it should stop supplying in the event of a power outage.

 

Current solution: I just checked the gas meter valve; it seems to be working fine.