Reliable oil/petroleum flaking

[sushieater]

I've spent something like 30 hours trying to make a reliable oil flaking boiler. My experience is that the flaking behavior is exceptionally flaky (pun intended). IME, flaking doesn't just suffer from load glitches, auto-saves can also screw up things. All flaking donors that I have tried besides chlorine have suffered catastrophic failures with randomly splattered crude oil or teleported petroleum (where it's not supposed to be) or flaking stopping to work, even though the donor is hot enough and the input oil is cold enough.

I've been able to make a reliable 5kg/s bead boiler (feeding 5010g oil beads). I've been unable to make a 2x 5010g bead drop work 100% reliably (2 5010g bead drops stacked on each other) in the same flaker/heat exchanger.

IME, the flaking chamber can randomly goes crazy and experience super heat-conduction. With a ceramic flaking tile, the boiler might randomly enter a state, where the output oil is well above 410C - which should be impossible (the conductivity is way too low).

I have a design that packet-izes the incoming oil to 5ks or more, which seems to work reliably with a chlorine donor. But given the insane amount of glitches I've experienced, I don't fully trust it.

[mathmanican]

I don't make 2x 5010g flakers either, if I'm using beads. A single bead is easy to make reliable. Where the resulting liquid flies is not 100% predictable, and can easily mess up another bead that is a few ticks away.  

Have you tried @Zarquan's approaches to flaking? Rather than making 5010g beads and flake from the side as a bead falls from above, instead flake from above and let your liquid flow horizontally. You can get 25kg/s out of your flaking with this approach. I'll let @Zarquan share his favorite versions (or you can search for some). If @Zarquan has seen issues, I'm sure we'll get comments, but I believe the approach is reliable. 

[wachunga]

The issues with flaking a double bead stream comes down to two things.

1) When you flake a crude bead, the resulting petroleum bead is in a different spot and you can be left with a transitory remnant crude bead.

2) Crude is the "densest" liquid and wants to be below all other liquids. When crude is directly above petroleum, chaos ensues.

In a single bead stream, you have 4 spaces between beads. There is plenty of room such that you don't get a crude above petroleum situation. In a double bead stream you can't space the beads to reliably avoid the crude above petroleum situation.

The stream on the right is baseline. The left stream gets flaked, a remnant crude bead exists for a single tick which is enough to stall the petroleum bead and throw it out of sync with the baseline. A double bead stream lacks the space to handle this stuff reliably.

IMO, if you are married to flaking beads and want more than 5 kg/s you should have 2 (or more) flakers. Join the outputs before going into your heat exchanger or build an exchanger for each flaking stream. I won't shut up about waterfall exchangers because they are super efficient and very easy to double up (or triple or quadruple or ...) like this.

I've never had any issues with single bead flaking boilers, granted that was only two or three times.

[sushieater]

On 10/6/2024 at 12:28 AM, mathmanican said:

Have you tried @Zarquan's approaches to flaking? Rather than making 5010g beads and flake from the side as a bead falls from above, instead flake from above and let your liquid flow horizontally.

Yes. I was maybe able to have that reliably working with chlorine. Everything else suffered catastrophic glitches with oil splattered around, the flaking stopping working or output petroleum suddenly becoming very hot (when loading or saving).

However, even the chlorine approach had issues, when oil flow was less than 5kg/s. The oil pressure in the flaking chamber was uncomfortably high (several hundred kg) and increased with a decreased oil flow. I was maybe able to fix the issue with packetizing the incoming oil to >=5k packets.

Given the amount of weird behavior I've experienced, I'm instead using an extremely well-behaved boiling chamber for my survival game. Consumption of molten iron from an iron volcano is 260g/s vs. 215g/s with the flaking design - which is good enough (but the flaking design was 2 tiles wider and thus had a slightly better heat exchanger). (The insulated outer shell is 21x21 tiles or 21x23 tiles.)

[Zarquan]

On 10/7/2024 at 11:00 AM, sushieater said:

Yes. I was maybe able to have that reliably working with chlorine. Everything else suffered catastrophic glitches with oil splattered around, the flaking stopping working or output petroleum suddenly becoming very hot (when loading or saving).

There isn't really a significant benefit to using chlorine anymore IMO.  I use ceramic tiles, which have the benefit of not trying to escape.

Does your horizontal boiler look like the one on the left side of this picture?