Factorio Forums

What did you do?

I am playing Pyanodon's (full suite). Relevant to this bug report, there is a building called "Coal powerplant MK 01". This building takes in a fluid (molten salt) and outputs another fluid (hot molten salt), burning certain types of fuel to do so. Three of those fuel types are coal, crushed coal, and coal dust.

I have a build where I have eight coal powerplants. I am converting raw coal into all three fuel types in secondary crushers, and direct inserting the fuel to the power plant.

What happened?
This works perfectly for seven of my eight power plants. However, one of my eight power plants (the third from the left, at location 222, 38 in the provided save) is also receiving some coal dust via a different belt. This particular building now has an issue where the fuel ends up in the output slots, as pictured below. If the machine continues running, more and more fuel will end up in the trash slots. I have seen this end up with literally dozens of trash slots taken up by fuel, as shown here.
What did you expect to happen instead? It might be obvious to you, but do it anyway!
I expect that fuel should not end up in the trash slots. I don't know what precise combination of factors leads to fuel being put in the trash slots, but I would expect that if a fuel slot is already occupied, an inserter would continue to report "destination full" rather than the fuel being moved to the trash slot and the inserter being allowed to insert fuel. Notably in this scenario, the *only* items being inserted are fuel (and the only item being removed is the spent fuel result, ash).

Does it happen always, once, or sometimes?
This always seems to happen, but the amount of time it takes to happen varies. This is possibly due to the fact multiple recipes with percentages are part of this setup (both the recipe creating the three fuel types and the ash processing recipe creating the coal dust have percentage outputs).

If you copy/paste all the buildings from the provided save somewhere else (such that the miners are on a raw coal patch), the same issue should occur in the new location given enough time.

Log file
<attached>
Save file
I recreated the issue in the map editor. This probably isn't strictly the minimum number of buildings/setup to reproduce the bug, but it's at least a pretty small file compared to a normal Py save. That editor save is attached.

Edit to add: This is similar to viewtopic.php?p=668646#p668646, where Rseding specifically asked for a new bug report with reproduction steps. To reproduce, copy/paste the setup in the attached save to a new location (where the miners are on a raw coal patch).

This sounds exactly like viewtopic.php?p=658425#p658425 with resolution being viewtopic.php?p=658523#p658523.

It looks like the machine at {222,38} was previously connected to a circuit network and as such it has control behavior and it has trash inventory present. As a workaround you can copy settings from other assemblers that are working correctly, this action will remove control behavior from this machine which will make inserters unable to insert into trash inventory and as such they will be blocking.

-- edit:

From what i can see, there is a fix for 2.1 that handles this exact case by making sure items are not moved to dump inventory (one owned by control behavior) when it is inactive (there are no wires present) or when there are circuit wires but "Set recipe" is disabled.

boskid wrote: Wed Apr 29, 2026 10:17 pm It looks like the machine at {222,38} was previously connected to a circuit network and as such it has control behavior and it has trash inventory present. As a workaround you can copy settings from other assemblers that are working correctly, this action will remove control behavior from this machine which will make inserters unable to insert into trash inventory and as such they will be blocking.

Well darn, you may be right.

I am 99% sure that I never connected the building to the circuit network until *after* noticing the issue and thus using circuits trying to fix it, but I cannot say this happened for certain (and I understand that I'm subject to confirmation bias - wanting to see the thing I suspect (a bug) be proven correct).

The copy/paste workaround did fix the building. I then tried to copy and paste the entire scenario into something more resembling the actual circumstances that occurred in my factory (shortages of molten salt, not-constant hot molten salt drain, actually consuming the pressurized steam as power, backing up and then releasing of ash, circuit-controlled inserters, etc.). I was hoping to find a specific set of circumstances that could cause the issue without ever connecting the building to the circuit network. While I'm not convinced that such a scenario doesn't exist, I nevertheless was unable to force the circumstance to happen. My best guess would simply be some combination of edge cases happening simultaneously (for example, internal fuel buffer running out, inserters hovering with multiple different items ready to insert, fluid ingredient shortage ending, and machine output no longer being full, all happening on the same tick), but I'm not sure.

This is something that has definitely happened to quite a few people, and I really am convinced that there is some combination of factors that can cause this to occur even without connecting the building to the circuit network, but I don't have the toolset (mentally) to test every possible scenario.

In any case, I'm glad to hear of the future resolution in 2.1! Thank you for your time looking into this.

Mod(s) can force the control behavior to exist regardless of the player connecting it to the circuit network. If they do that, this issue will start happening.

It's *partially* fixed for 2.1. I see another potential issue that the 2.1 fix doesn't address however I can add another check that should fully fix it (and cause inserters to block until the fuel inventory empties out and allows the item(s) to be burnt or input to the machine).