It also sounds like these bases were built without upgradability in mind. Larger separation of concerns and modularization of areas would facilitate this piecemeal approach to upgrades a lot more easily. Then you would be looking at rebuilding individual modules rather than rebuilding a whole ...

Your space research center can only reach about 70k at most, and that's your limit.


You can build more than one


If I use only belts and inserters, the maximum is around 70k


Something about your math isn't mathing. Legendary stack inserters do 120 items per second in chest to chest ...

Ok, sure, if you hyper-optimize it,


Even the setup you posted can support 24k raw spm before productivity and biolab bonuses. And you can still add bots on top of that. Or if you really really don't want to use bots nothing is stopping you from putting unlimited numbers of regular labs on space ...

airbrush wrote: Fri Jun 12, 2026 11:31 am The root cause

To give the short version I've done a half dozen times before, inserters alone let you sustain 36k raw spm. Iirc your physical hardware will bottleneck before the landing pad does.

airbrush wrote: Fri Jun 12, 2026 4:36 am and there is a bottleneck

Even without using bots, you can get 50% more throughput by going from pad -> chest -> two belts (add extra chest->chest steps for more space)

This discussion and others show how in nearly every case, it's better to use productivity modules.

However, if you have dedicated "quality science" assemblers that use leftover quality materials from your base, I guess another solution could be just having separate labs for each quality.


For ...

Yep. Ever since discovering this I usually just don't bother putting more than one shield module for rocks on tanks. Better to use the space and power for exoskeletons to get hit less. Or for lasers/robots.

I'm not understanding how this setup works, then.


It's abuse of the enemy AI. If they can path to a turret they prefer doing that over attacking walls, so they enter the fatal funnel to approach the flamethrower turret. While doing so the flamethrower turret coats the whole corridor in fire ...

macdjord wrote: Tue Jun 09, 2026 8:27 pm But why? You want the biters attacking your walls, because walls have lots of HP and are cheap to replace if they die.

Because with the setup I described, they usually hit absolutely nothing.

Higher HP obviously means the turret will last longer while under attack. With gun and laser turrets, this can be very useful,


There are strategies that involve sticking flamethrower turrets in a narrow pocket but in front of the wall to encourage the biters to ignore the walls. This is usually ...

I'm quite aware of how it works. And the way it works causes some setups to break when flipped . This is fundamentally not expected behavior - ergo, a bug.


Don't change the chirality of setups that rely on chirality. There's no way to do what you want without breaking other things or adding ...

There's some interesting logistics bottlenecks like ingots that are so much more efficient to ship than plates that it complicates things, but once you start getting the really cheap logistics options the savings from beaconed productivity setups go wild.

I think the unsatisfying answer for SE is almost always "depends on your tech levels"

boskid wrote: Mon May 11, 2026 7:37 am if there would be some stray characters nearby or biters they should be also grinded and output onto belts

Sounds like a fun mod idea. Turn off resource patch restrictions and put biter drills in wall choke points. Outputs spoilage.

Try this wiki page:

https://wiki.factorio.com/Inserters

It has a ton of data and tables that you can use to get all the information you need for every inserter at every research level and every quality.


https://wiki.factorio.com/inserters#Inserter_throughput Specifically this subheading. OP ...

Differing quality of ammunition can't be stacked, so it will indeed run out of ammunition while the switch happens because the inserter can't start work until the turret is empty.

I started a more in depth reply on break points, but realized I don't have enough info on how lightning works beyond "about one strike per chunk per 600 ticks". A lot of it comes down to the probability distribution and timing of strikes and other than averages I haven't found much to work with.

My ...

I cannot remember the last time i paid anywhere close to $90 for a computer game, its just out of touch.


With microtransactions the expected lifetime value of a customer is much higher than $90CAD. Heck, a lot of them get more than Factorio out of their players the first time they swipe ...

Looks like a classic "sorting alphabetically instead of numerically when able"

What shall i say?


Probably start with what the graph is? 😅


The catch is that you have more electricity storage capacity with standard collectors, so when the night is over you're running from the stored energy in your collectors for a longer time.


Any collector runs dry on its own in at ...