Fast Acid-Neutralisation and Steam-Condensation, throughput issue?

[JackTheSpades]

I'm trying to design a megabase but I've run into somewhat of an issue on Vlucanus.
For all the water I need to handle my oil processing, the calculator shows that two legendary Cryoplants powered by a legendary beacon with legendary speed modules would be enough.
And I suppose in raw numbers it actually would be but the problem is that the steam doesn't really move from the Acid-Neutralisation (AN) to the Steam-Condensation (SC). The AN is full with 10k of steam but the SC is just slowly filling up with the 1k steam needed to start crafting. Using more SC units just makes them all slow each other down with the throughput even though the AN is still full. Pumps are no help either, regardless of how many I use.

Notice that the pipe is nearly empty.

One to Many

grafik.png (1.27 MiB) Viewed 544 times

I've tried some more extreme setups for testing which at least increases the water output but it still only runs a fraction of the time.

Direct Steam to Water setup

grafik.png (378.61 KiB) Viewed 544 times

Is my only possible solution to use multiple AN directly to SC units to get the water I need?

[Tertius]

Add one or more storage tanks to the steam pipe as buffer. The steam is generated in so huge quantity with so much throughput, it's too much to hold within just a pipe. The location of the tanks doesn't matter, since everything is just one big volume, just somewhere connected to the steam.

[JackTheSpades]

A tank makes no difference since the problem seems to be that the Condensation doesn't fill fast enough.
The tank here holds 12k steam. If I use pumps I can get the fluid content higher (around 20k) but that doesn't make the SC work continuously either.
I can get the same result by using pumps directly into the SC without any tanks too.

Steam fill to slow

Water.gif (5.25 MiB) Viewed 488 times

[eugenekay]

129166 is possibly related.

Pumps should not be necessary over such a small distance - directly connecting the Pipes will let them flow Instantly. Just make sure that all your Machines Input/Output are connected to a Pipe; that there is enough Tank buffer in the Fluid networks; and don’t try to use a full set of Speed modules.

Good Luck!

[Tertius]

Now I think deeper, I remember the fff about the new fluid system, which says the maximum fluid throughput of a single fluidbox is 100 per tick or 6000/s.
If you use each of the 3 inputs/outputs, you get a maximum of 3*6000/s=18000/s. You also see the reason why this machine got 3 inputs and 3 outputs for the same fluid.

However, actually usable is just about 4200/s, which is also described in the wiki: https://wiki.factorio.com/Acid_neutralisation
So the maximum usable speed of a cryogenic plant with acid neutralization is about 6.3. The fluid system simply doesn't drain the output of the recipe fast enough.

A setup like this will work, with about these as ratios:
The beacon is a normal beacon with a speed module 1, which results in a speed of 6.5 for a legendary plant.
The 7 condensers output 3150/s water. You could boost them slightly to get less plants, but not very much.
I would say don't use legendary plants, since you cannot profit from their inherent speed bonus. Much too expensive. Normal plants with a few speed modules or instead of this one or two speed beacons are sufficient to reach the limit. Connect all three inputs and outputs!

07-29-2025, 00-53-49.png (756.44 KiB) Viewed 445 times

[JackTheSpades]

You are definitely right that using legendary cryo plants is a waste.
But I think the actual speed limit seems to be 9.5.

I did a brute force setup in which I just checked how much I can push the speed with the condensation plant working continuously and this is what I found. I used a latch to check if the working signal ever disappeared.

grafik.png (504.04 KiB) Viewed 391 times

It uses a direct 1:1 insert from the acid neutralization and runs at 855 water/second. Your setup has 7*180=1260 water/second but requires more space, thought definitely friendlier on the module usage.
But given how much water I need the space efficiency of the 1:1 setup is probably worth it.

[Tertius]

By using that many plants, I just mirrored your first screenshot. Since given the correct amount of speed boosts one water condensing plant is able to fully consume the output of one acid neutralization plant, the most space efficient setup is the direct connect. However, there are subtle and surprising (but small) differences.

Acid neutralization plant speed = 6, water condensing plant speed = 12. That's exactly the steam ratio between the 2 recipes. Maximum output for these speeds is 1080/s water. Real output:

07-29-2025, 12-50-58.png (1.36 MiB) Viewed 351 times

The water tank is just to smooth output rates for analysis. The output rates vary by +-10. The screenshot contains the pump averages.

The highest output is the top right variant. Not the bottom left as one would expect.

[JackTheSpades]

I hadn't considered using pumps to measure the output. So I based my design only on the idea of what design works continuously without ever triggering the "Fluid ingredient shortage". Mainly because all of the calculation tools I use expect the machines to run like that to produce their expected output.
But even with the flickering work the average output of 1k water/sec is still much better.

For the sake of testing, I've tried putting two of the "one pipe in between" configurations next to each other such that the pipes connect but it doesn't seem to really hinder the output, which still runs at 2*1035-ish.

grafik.png (557.36 KiB) Viewed 329 times

Even with three combined, the ratio holds.

grafik.png (792.73 KiB) Viewed 329 times

Funny enough, with one big connected steam pipe the output plummets to 2.5k which is effectively 16% loss.
I guess that as the pipe fills up the output of the AN units slows down which in turn hampers the whole system.

grafik.png (723.3 KiB) Viewed 329 times

[jdrexler75]

The highest output is the top right variant. Not the bottom left as one would expect.

I think it would be interesting to try this in 2.0.47 or earlier, because of the apparent regression of direct fluid connections since 2.0.48.

[Tertius]

There's really a difference with the direct connection in 2.0.47. Every other build is the same throughput, but with direct connection in 2.0.47 water output can go up to ~1974/s, which corresponds to a speed of ~11 / ~22, which is steam throughput of 22000/s, which means a throughput of 7333/s per output, which is over the maximum of 100 per tick. I'm unable to tell if an unintended behavior has been fixed or introduced.

07-29-2025, 15-38-09.png (412.79 KiB) Viewed 292 times

No save file required, just use the blueprint in 2.0.60 and in 2.0.47.

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

Copy blueprint

[jdrexler75]

I'm unable to tell if an unintended behavior has been fixed or introduced.

Since the ticket has remained open for two months now, I suspect it wasn't an intentionally introduced change, so I bet there will be some corrections though how much (and in which direction) is anyone's guess...

But right now is probably not the best time to finalize the design of a large number of high-throughput fluid connections.