Ethanol disappearing from turbine room?

[myxal]

I'm trying to cool turbines in an LOX/LH2 machine to some reasonable temps without a separate cooling loop or heat injector. Here's the build:

I start with 2kg/cell at the turbines, but after ~5 cycles it's almost all gone - gas cells are a mere 50 g/cell, an other than those 3 liquid cells on the elevated floor, there's no liquid at the turbines. No solid ethanol around, either (although it periodically appears at the right turbine's upper left corner, before immediately melting and boiling).

Another cycle later there's only gas cells at milligram pressure, and some precipitation that doesn't form beads upon reaching the floor.

 

EDIT: Hmm, reducing the heat transfer from 6 granite tiles + 2 gold TSPs to 3 granite tile and 1 granite TSP seems to have fixed the issue (mass deletion and ethanol freezning). Though I'm still curious to find out why it happened. Some funky deletion when dealing with solid debris, liquid precipitation, and gas atmosphere in the same cell?

 

[MikasaTanikawa]

You periodically have completely dry floor in turbine room. Every time when gas condensates, falls on ground and creates less than 10g of liquid in tile, game will delete it.

[myxal]

@MikasaTanikawaThanks, so this is related to precipitation -> atmosphere change.

I now do remember that trying to get <10g liquid blob out of a liquid vent was a futile effort. Feels weird that the blob exists for 2-3 ticks , potentially even displacing other liquids before just disappearing Also, "floor" isn't restricted to solid cells, landing on denser liquid will also delete it.

So I suppose the possible solutions are:

• increase mass of ethanol so that more mass is condensing at any moment
• reduce the width of the cold body so that the gas is condensing in 1 column