Polluted water bubbler

[Zarquan]

I was thinking back on one of my old builds, where I used an infinite waterfall to compress polluted water to get massive amounts of polluted oxygen out of one tile of water.  However, the physics has changed, so I thought about how to build a revised model.  

Spoiler

For those who do not know, every game tick (which is 5 times a second), a surface tile of polluted water has a 0.1% chance of discharging 0.1% of its mass as polluted oxygen at a 1:1 ratio.  For example, if you have 1000 kg of polluted water and it bubbles, you end up with 999 kg polluted water remaining and 1 kg polluted oxygen.  Or, if you have 100000 kg of polluted water, you end up with 99900 kg of polluted water and 100 kg polluted oxygen.  This build maximizes the amount of water in a single tile while still letting it offgas.

The polluted water is compressed using a variant of the infinite waterfall compressor.  The polluted water is compressed on to a single tile under an airflow tile.  The waterfall is made with a 500 g packet of CO2 to the right of the water and variable amounts of polluted oxygen.  The structure on the left with the clean water is a bead pump with an infinite waterfall powering it.  The effect of the bead pump is to move gas from the chamber with the atmosensor to the chamber above, as polluted water does not offgas if the gas pressure is too high (>1800).  The atmosensor is the only required automation, and it is set to above 5 grams pressure.  The room with the polluted water with the vents acts as a water source.

When the polluted water off-gasses, it releases a massive amount of polluted oxygen.  This causes the atmosensor to be triggered and the door to open, exposing the polluted oxygen to the bead pump.  Since bead pumps pump proportionally, it quickly decompresses the room from around 200 kg of pressure to less than 1800 g in around 20 seconds.  This is much better than the performance of my door pump setup in the original, moving the polluted oxygen up to the upper chamber.

Theoretically, this should output polluted oxygen as much polluted oxygen as you input polluted water.  This is because the average amount of water inputted would be fairly constant and the polluted oxygen emitted is proportional to the water currently in the tile.  So, there will be a value the polluted water will trend towards depending on the amount of polluted water poured in to the system.

The probability of the polluted water not bubbling in a cycle is around 5%, so this should be fairly regular.

DISCLAIMER:  This prototype was built in debug mode, but creating it in game would not be complicated, except for collecting the mass of water I have here. 

Spoiler

This is a link to my original build.

WARNING:  THE BUILD IN THE LINK NO LONGER WORKS!