Factorio Forums

SoShootMe wrote: Wed Sep 07, 2022 11:15 pm Nuclear power without the Kovarex enrichment process is (as you've proved) possible, but I find not worthwhile in practice.

I wouldn't try running a large nuclear power plant without Kovarex, but you can quite happily run a 4-core one (ample power for everything up to a rocket launch) on natural U-235. Kovarex is an expensive tech, and I'd rather not wait for it before turning on the switch that solves all my power problems for the foreseeable future. In my current game (16 hours in) I turned on nuclear power 2 hours ago, I have just discovered Kovarex, and despite having been burning some, I had exactly the right number of U-235 on hand to start the enrichment cycle; if I'd waited to turn the reactor on, I doubt I could even afford to research Kovarex yet. The main risk with doing it this way is not that you won't have enough U-235 to start Kovarex, but that during the ensuing Kovarex build-up the enriching centrifuges hog all of it, not allowing any through for the reactor.

ColonelSandersLite wrote: Wed Sep 07, 2022 7:56 pm

eBagger wrote: Wed Sep 07, 2022 8:54 amAnd by the time you get to nuclear you should have fields and fields of solar panels like I do, so it’ll take quite a few reactors to run your factory.

Here's the thing from my perspective -
Solar is pretty expensive, takes up a ton of space, and is very boring. Sooooooo boring.

I will only build it in exactly 3 situations.

1 self sufficient off the grid remote area radar outposts that are just there to reveal the map. These get torn out once they have done their job.

2 I'm uranium poor for some reason and I need more power to get more uranium.

3 Megabase territory for the UPS.


Don't ignore the cost of solar panels. They *aren't* cheap once you multiply their base cost by "fields and fields" and it takes resources to clear the land to build them.


Did I mention they're boring?

I got so incredibly bored with them that I have abandoned the tilable designs for something a bit more interesting to look at.



celtic power A.jpg

celtic power B.jpg

So... I guess I'm running on celtic power now. That's green right?

Thats awesome

I dunno I find giant fields of boring square solar panels appealing for some reason. I suppose I see it as a testament to how far I’ve come from when a small square of a few hundred solar panels and accumulators can cover my entire bases powers needs.

I like how active these forums are. I expect to sign in and maybe see a reply or two, and instead seeing several full discussions on the matter.

I see what you guys are saying, again its like I was thinking and I just don’t have the experience of seeing there benefits yet. I dunno we’ll see, launching my first rocket is my priority at the moment, and I’m only a few researches away so woohoo. I’ll experiment more with nuclear expansion after I launch and attempt some city-block style factory organization.

Khagan wrote: Thu Sep 08, 2022 1:44 am

SoShootMe wrote: Wed Sep 07, 2022 11:15 pm Nuclear power without the Kovarex enrichment process is (as you've proved) possible, but I find not worthwhile in practice.

I wouldn't try running a large nuclear power plant without Kovarex, but you can quite happily run a 4-core one (ample power for everything up to a rocket launch) on natural U-235. Kovarex is an expensive tech, and I'd rather not wait for it before turning on the switch that solves all my power problems for the foreseeable future.

Yes, Kovarex is an expensive technology. For context, it's around double the total cost of researching Nuclear power and building a typical 2x2 nuclear power plant and the centrifuges to provide enough U235 (without Kovarex) for it. Obviously I have a different play style but I guess it amounts to the fact I'm not desperate for more power so it's no hardship to wait longer. In fact if my science output was high enough to need nuclear power I'd probably be able to research Kovarex before I had produced the first 40 U235 to make use of it. I probably hang on to steel furnaces too long; an earlier switch to nuclear power (before researching Kovarex) and electric furnaces would save pollution. I'll have to try it.

SoShootMe wrote: Thu Sep 08, 2022 8:12 am I probably hang on to steel furnaces too long; an earlier switch to nuclear power (before researching Kovarex) and electric furnaces would save pollution.

It doesn't have to be either/or. I don't start using electric furnaces until I have nuclear power up and going, but that doesn't mean I scrap my old steel ones in a hurry. The new electrics are additional to the steel, and both sets get scrapped (or reused) together once I start building heavily moduled and beaconed smelters for the infinite game.

Khagan wrote: Thu Sep 08, 2022 8:48 am

SoShootMe wrote: Thu Sep 08, 2022 8:12 am I probably hang on to steel furnaces too long; an earlier switch to nuclear power (before researching Kovarex) and electric furnaces would save pollution.

It doesn't have to be either/or. I don't start using electric furnaces until I have nuclear power up and going, but that doesn't mean I scrap my old steel ones in a hurry. The new electrics are additional to the steel, and both sets get scrapped (or reused) together once I start building heavily moduled and beaconed smelters for the infinite game.

Do electric furnaces produce less pollution or why would you swap? They are the same speed as steel furnaces and so far I always assumed switching to them is pointless unless you add modules. And since modules are best added from the top of the production chain down the furnaces are the second last place to add modules, just before adding them to miners (which is currently a hot debate in another thread that compares it with the mining bonus research). I tend to have nuclear power before I amass module 3s to put into smelters.

It kind of makes sense to build electric furnaces as soon as you can because any steel furnace you build is eventually wasted when you do replace them later (compared to electric furnaces with productivity modules scrapping the steel furnaces makes sense). But till you have the modules the old steel furnaces perform just as well as electric.

I guess one could start early with level 1 modules?

Edited - disregard this post. had a mistake where I read some numbers from the wrong save.

How many centrifuges do you have refining uranium? Maybe I just would need more to get enough light green uranium to even run the trains on it pre Covarex.

mrvn wrote: Thu Sep 08, 2022 9:13 am Do electric furnaces produce less pollution or why would you swap?

Assuming continuous running, a steel furnace produces 4 pollution per minute (p/m), while an electric furnace produces only 1p/m. Sounds great but a steel furnace also creates ~0.5p/m mining coal to power it, while the electricity for an electric furnace creates ~4p/m assuming coal miners -> boilers -> steam engines.

Overall that means electric furnaces produce around 10% more pollution than steel furnaces, using coal power. With solar or nuclear power, the latter becomes basically 0p/m, making electric furnaces a big improvement. The benefit is offset by one-off costs in both cases (both for power generation and the furnaces themselves), and a small running cost for nuclear.

You can also use 2x efficiency module 2 in both miners and electric furnaces, in which case the pollution from each furnace reduces to 0.2p/m plus ~0.75p/m for electricity, similar in total to not having modules but using solar or nuclear power. Again there is a one-off cost to produce the modules, but I'm not sure how that compares to building solar or nuclear (instinct tells me not well, but I could be wrong).

I'm gonna kind of group some of these general thoughts together so as to avoid multiple posts. Not trying to misrepresent anyone's point or anything.

mrvn wrote: Wed Sep 07, 2022 10:35 pm I just find it uggly. And then you always design the chest field to small and at some point it's full and you only notice because you have a blackout.

Well, at least in my case, startup bases are always kinda ugly. I can't say that I'm super concerned with aesthetics until I push artillery range upgrades out there and establish a new perimeter that encompasses my main base and my outposts too. Probably a death world habit as much as anything though.

Honestly though, the u-238 buffer size isn't as big of a deal as you think it is. Like I said, I ran nuclear power with just 50 chests for 82.5 hours. I was also running nuclear trains for 80 of those hours. When I finally did get around to building the kovarex line, I had consumed 2.4k u-235 and still had 750ish in the bank. That's enough to run for at least another 20 hours without any more uranium production. Also noteworthy - at this stage I had not yet set up my reactors with any sort of smart logic either so they where just on continuous burn the whole time.

mrvn wrote: Wed Sep 07, 2022 10:35 pm I never looked into uranium bullets, maybe that does eat up a good part of the uranium. Way before that I switch to lasers.

See, I never use laser turrets. There's 3 main reasons.

1 - My "main" defense turret is the flame turret. No question. The choice between laser and gun turrets, for me, is only really as it pertains to protecting the flamers. Since I want flamers, for things to be self repairing, and eventually artillery shells too, there has to already a logistics system in place to do those things so it's not really an issue to just send bullets along that same logistical route.

2 - I haven't done the actual math since 0.18.something so it's possible it has changed, but at least as of then - Gun turrets with *red* ammo have a higher DPS at all tech levels vs all biters with the sole exception of behemoths just before space science. Uranium mags are a whole new ballgame. They put out 3-4 times the DPS of laser turret with tech parity. iirc, with a *whole lot* of infinite tech research lasers will eventually outpace guns but by that time biters are just instantly dying when a turret looks their way anyways so it's kind of a moot point.

3 - On paper, laser turrets are definitely the better choice for turret creeping but, IMHO, turret creeping isn't even particularly relevant once they're available. If you have chem science, you can have a tank with HE shells and with just a bit of skill, you can just butcher the biters. In the end game, spidertrons and artillery make it wholly irrelevant.

There are other reasons too, but those are the biggies.

SoShootMe wrote: Wed Sep 07, 2022 11:15 pmStoring the excess U238 is a pain but you also need a reasonable buffer of fuel cells to ensure that "on average" qualification doesn't cause you to run out, which in turn means you need to have produced a good handful of U235 before you can rely on nuclear power.

Nah, you can just fire that thing up pretty much immediately. That's what I do. You have to keep in mind that we're talking about transitioning from coal to nukes here. If there's a small interrupt at, lets just say refueling cycle 3, before the 235 starts to accumulate, it's not gonna kill you or anything. You'll just spend a few minutes at the same power level you where at a few minutes ago. The 235 will pile up in short order so it's not going to be a hazard past the first half a dozen or so refueling operations as long as you're not underproducing.

SoShootMe wrote: Wed Sep 07, 2022 11:15 pmYou could instead have put that U235 towards getting enrichment running sooner, guaranteeing fuel cell supply and (practically speaking) enabling the other uses of U235, nuclear fuel and atomic bombs.

I run nuclear trains without kovarex too. See the numbers above. The only one of those things I would worry about is nukes. Nukes do eat uranium up pretty quickly. You probably *do* need to set up kovarex before you can really use them. There isn't really a rush though. It's a 6 pack science and I'm more talking about the blue science stage. Not a big fan of the shoulder fired nuke launcher myself. I like to lob M.I.R.V.s around but that's endgame stuff.

The other thing I would watch out for is nuclear train fuel on belts. I'm a belt everything kinda guy but I've changed my mind on nuclear train fuel specifically. A single side of a belt of it running to and through just my train stations would probably hold several hundred, if not thousands, of fuel rods that just aren't going to get used.

mrvn wrote: Thu Sep 08, 2022 12:02 am With covarex you could even use it in furnaces.

Alright now, that's just excessive. It's just a short hop away from using it in boilers to power regular ol' steam engines.

mrvn wrote: Thu Sep 08, 2022 9:13 am

Khagan wrote: Thu Sep 08, 2022 8:48 am

SoShootMe wrote: Thu Sep 08, 2022 8:12 am I probably hang on to steel furnaces too long; an earlier switch to nuclear power (before researching Kovarex) and electric furnaces would save pollution.

It doesn't have to be either/or. I don't start using electric furnaces until I have nuclear power up and going, but that doesn't mean I scrap my old steel ones in a hurry.

Do electric furnaces produce less pollution or why would you swap?

Last time I checked, when you look at the whole system, electric furnaces produce slightly more pollution than while you're still on boilers. Not a huge amount more, just a bit. They will produce less pollution when you're on either solar or nuclear power though. There's another thing to consider - electric furnaces have an idle power draw but steel furnaces do not.

Personally, I'm with Khagan on that one. I'm never in any kind of a rush to rip out steel furnaces and replace them with electrics *unless* there's a serious fuel shortage on the map or something. The fact that you don't have to route in coal to feed electric furnaces is nice but your existing smelting lines already have a fuel source. I usually don't make a switch until I want to start doing some offsite smelting. If the goal is just to launch a single rocket, that's never.

mrvn wrote: Thu Sep 08, 2022 12:39 pm How many centrifuges do you have refining uranium? Maybe I just would need more to get enough light green uranium to even run the trains on it pre Covarex.

Not 100% sure how many I had when I fired up the first reactor but I quickly went to 56. Not so much to get u-235 quickly but in order to make a lot of bullets quickly.

That should work out to a u-235 every 30.6 seconds.

I just want to reiterate though that u-238 storage in the form of bullets is probably not insignificant. If you're not going to consume u-238 on bullets, it may well be wise to set up kovarex much more quickly than I did here. I wouldn't stress out over getting the first 40 as a lot of people do though.

mmmPI wrote: Wed Sep 07, 2022 9:48 pmThat's good to hear you plan to share

eBagger wrote: Thu Sep 08, 2022 2:20 am Thats awesome

SoShootMe wrote: Wed Sep 07, 2022 11:15 pm Very impressive.

mrvn wrote: Wed Sep 07, 2022 10:10 pm Wow, blueprints please.

I'm glad you guys enjoyed that. I'll share the tools I made once I have them in a more user friendly state. ATM, they're just kinda kludgy.

mrvn wrote: Wed Sep 07, 2022 10:10 pm Did you make them by hand or did you write a image-to-blueprint generator?

Those questions where answered in my last post just before you asked. A bit of both.

ColonelSandersLite wrote: Thu Sep 08, 2022 2:27 pm Not 100% sure how many I had when I fired up the first reactor but I quickly went to 56. Not so much to get u-235 quickly but in order to make a lot of bullets quickly.

That should work out to a u-235 every 30.6 seconds.

I just want to reiterate though that u-238 storage in the form of bullets is probably not insignificant. If you're not going to consume u-238 on bullets, it may well be wise to set up kovarex much more quickly than I did here. I wouldn't stress out over getting the first 40 as a lot of people do though.

That totally explains it. I usually set up one or two as soon as I research centrifuges and let it churn away on uranium. Then by the time I have Covarex I likely have the 40 U-235 needed.

I don't stress about it but I do care about it. Worst case when I get Covarex I just have to wait a bit more before the reactor starts up. But it's annoying if you get the research, get all fired up and build the centrifuge(s) and then have no U-235 to fire it up and test it.

What I do recommend though is to set up one filter inserter taking U-235 and putting it in a chest with "U-235 < 40", potentially 80. You can put it behind a splitter so it only takes every second U235 if you like. Even if it saves only 20 minutes like in your case.

ColonelSandersLite wrote: Thu Sep 08, 2022 2:27 pm

SoShootMe wrote: Wed Sep 07, 2022 11:15 pmStoring the excess U238 is a pain but you also need a reasonable buffer of fuel cells to ensure that "on average" qualification doesn't cause you to run out, which in turn means you need to have produced a good handful of U235 before you can rely on nuclear power.

Nah, you can just fire that thing up pretty much immediately. That's what I do. You have to keep in mind that we're talking about transitioning from coal to nukes here.

"Transitioning" is not the same as "rely[ing] on". I think we basically agree, except I don't see the point of starting the reactor when it might run out of fuel, because that means I can't (yet) rely on nuclear power. By the time I can rely on it, I'm closer to researching Kovarex, and it feels more worthwhile to delay a bit longer, so I keep the U235 and get Kovarex running sooner.

As well as constraining the amount of U238 buffer required, it also neatly avoids the potential issue of starving fuel cell production when you later get Kovarex going, which is also an "it might run out of fuel" situation.

Maybe I'm overcautious (definitely). Maybe I got hurt by consequences of low power as a child .

ColonelSandersLite wrote: Thu Sep 08, 2022 2:27 pm

mrvn wrote: Thu Sep 08, 2022 12:39 pm How many centrifuges do you have refining uranium?

Not 100% sure how many I had when I fired up the first reactor but I quickly went to 56. Not so much to get u-235 quickly but in order to make a lot of bullets quickly.

Ah, yeah, with that many centrifuges, U235 isn't going to be a problem, assuming the U238 has somewhere to go. Must be a very large uranium patch (or multiple patches) though.

mrvn wrote: Thu Sep 08, 2022 4:22 pm That totally explains it. I usually set up one or two as soon as I research centrifuges and let it churn away on uranium. Then by the time I have Covarex I likely have the 40 U-235 needed.

Well you see, one centrifuge works out to (with a degree of randomness) approximately 40 u-235 in 19 hours. I'm definitely not going to wait 10 or 20 hours to get my nuclear plant going. Consider instead building 20 centrifuges so you have all 40 in under an hour so you can fire that bad boy up. If that's too many centrifuges for your needs once you have kovarex, you can always repurpose some of them into kovarex, recycling, and train fuel.

That's why I say don't stress it. Just start that reactor and if you don't have enough u-235 to start kovarex - no big deal. You'll have it in less than an hour.

SoShootMe wrote: Thu Sep 08, 2022 5:25 pm Must be a very large uranium patch (or multiple patches) though.

Pretty good size patch, yeah. I'm basically running rail world with standard biter settings re-enabled right now. Even a fairly lousy starter patch with straight default settings should be able to run about 20 centrifuges though and a moderate expansion patch on default settings should be about 40+.

20 Centrifuges is total overkill, especially with beacons. Even without you can easily run a 2x6 with a pair of centrifuges (one refining, one covarex). Add beacons around it and you soon drown in processed uranium.

One Covarex centrifuge with 2x productivity modules 3 surrounded by beacons produces an U-238 every ~8 seconds.

I'd like to add my version of uranium handling.
As soon as I found an uranium ore patch somewhere and also researched uranium processing, I start mining uranium ore and collect it in my base.
I build an array of 13 centrifuges to refine ore to U238 and U235. To speed up the bootstrap, I put speed modules into the centrifuges if I have some, and even beacons with speed modules.
Both U238 and U235 are stored in chests. For U238, there are 2 chests reserved. This is usually enough space for U238 until 40 U235 have been refined.
Usually, kovarex is researched before 40 U235 items have been collected.
As soon as kovarex is researched, whether I have 40 U235 or not, I build the remaining part of my uranium factory. I put the U235 collected so far into the kovarex centrifuge, then it's collecting the remaining U235 pieces by itself and auto starts the loop as soon as 40 pieces are in the centrifuge.

It's all automatic: no U235 is consumed for fuel cells or nuclear fuel as long as the kovarex loop isn't overflowing. That means 40 (or was it 80?) more U235 pieces in the centrifuge circle from the process itself, then it overflows and the overflow is getting to the assembling machine/centrifuge that is producing fuel cells and nuclear fuel.

As soon as kovarex is overflowing and some fuel cells have been produced, the beacons are removed and the speed modules are all replaced with production modules. First, all 13 centrifuges are used for getting the first U235, later they are repurposed for kovarex, reprocessing and nuclear fuel as needed.

There is also a limiter that checks for U238 storage. If the storage gets full (> 500), it stops the ore supply belt to the centrifuges, so there is only produced some if there is demand. Of course this limiter becomes only active after the first kovarex circle was detected. I count the items pulled from the kovarex centrifuge. If U235 > 1, the U238 limiter becomes active, because from then the relevant U235 source is the kovarex centrifuge, not the regular centrifuges any more.

There are also 2 centrifuges reserved for reprocessing of used up uranium fuel cells. They feed into the U238 loop, and they never congest, because they are not subject to the 238 limiter.

It's all not exactly tileable, but it's scalable: there's enough space for up to 4 kovarex centrifuges, 2 reprocessing centrifuges and as many regular refining centrifuges as you like. If I need even more production speed, I add beacons with speed modules. It's all in one row, so the usual beacon rows fit nicely.

ColonelSandersLite wrote: Thu Sep 08, 2022 10:05 pm

mrvn wrote: Thu Sep 08, 2022 4:22 pm That totally explains it. I usually set up one or two as soon as I research centrifuges and let it churn away on uranium. Then by the time I have Covarex I likely have the 40 U-235 needed.

Well you see, one centrifuge works out to (with a degree of randomness) approximately 40 u-235 in 19 hours. I'm definitely not going to wait 10 or 20 hours to get my nuclear plant going. Consider instead building 20 centrifuges so you have all 40 in under an hour so you can fire that bad boy up.

mrvn wrote: Thu Sep 08, 2022 10:30 pm 20 Centrifuges is total overkill, especially with beacons. Even without you can easily run a 2x6 with a pair of centrifuges (one refining, one covarex). Add beacons around it and you soon drown in processed uranium.

I put down 6 (with internal PM1 but no beacons) as soon as I start mining uranium. That gives me enough U-235 to start running a 4-core reactor after a couple of hours. Then (since power is no longer an issue) I add beacons with SM1 which make the existing centrifuges run at speed around 2.5 (so it's as if I had 15); meanwhile I'll research Kovarex so that the tech is ready at about the same time I have enough U-235 to start it (another couple of hours). I then build up to 4 centrifuges running Kovarex (again with internal PM1 and beaconed SM1); by which time, as mrvn says, I'm drowning in the green stuff.

Tertius wrote: Thu Sep 08, 2022 11:51 pm It's all automatic: no U235 is consumed for fuel cells or nuclear fuel as long as the kovarex loop isn't overflowing. That means 40 (or was it 80?) more U235 pieces in the centrifuge circle from the process itself, then it overflows and the overflow is getting to the assembling machine/centrifuge that is producing fuel cells and nuclear fuel.

I think it could be up to 120 uranium235 that inserter fit in a centrifuge at the same time, if you consider 40 that are currently being processed, the inserter non-controlled by circuit will still fill up to 80 in the centrifuge. it will lower down to 40 when a new cycle start, and the inserter will add 40 back to make it 80 again while there would be 40 that are stored in a progress bar at all time. From my observations.

It could of course be made better with circuit or belt magic so that the first centrifuge doesn't stockpile uranium 235 and let it be available for a second machine. Which is a nice challenge

There is also a limiter that checks for U238 storage. If the storage gets full (> 500), it stops the ore supply belt to the centrifuges, so there is only produced some if there is demand. Of course this limiter becomes only active after the first kovarex circle was detected. I count the items pulled from the kovarex centrifuge. If U235 > 1, the U238 limiter becomes active, because from then the relevant U235 source is the kovarex centrifuge, not the regular centrifuges any more.

Interesting approach, my personnal go to is to output the U235 from regular centrifuge as the main source of U235 for power, and then when it's not enough, taking some U235 from kovarex centrifuge using splitter priority input on 2 belts.

This means kovarex process backs up at some point because there is nowhere to ouput the U235, i'm not touching its input.

I then have a storage/buffer chest between the kovarex's centrifuge and the priority splitter from which inserter can take U235 for train or ammo if there is enough reserve for power safety.

Tertius wrote: Thu Sep 08, 2022 11:51 pm It's all automatic: no U235 is consumed for fuel cells or nuclear fuel as long as the kovarex loop isn't overflowing. That means 40 (or was it 80?) more U235 pieces in the centrifuge circle from the process itself, then it overflows and the overflow is getting to the assembling machine/centrifuge that is producing fuel cells and nuclear fuel.

It takes 40 to start the cycle and 2 cycles worth of inventory can be loaded while that runs. So the limit should be 120. But if you have 79 in the centrifuge and a stack inserter then it can pick up 12 more and insert them for a total of 40 + 79 + 12 = 131. If you have multiple inserters it can be even more.

Since you start with 40 and the recipe runs 60s a plain centrifuge then needs 91 minutes to produce the first output that isn't put back.

Speed modules or multiple centrifuges are really speeding things up.

mmmPI wrote: Fri Sep 09, 2022 7:07 am It could of course be made better with circuit or belt magic so that the first centrifuge doesn't stockpile uranium 235 and let it be available for a second machine. Which is a nice challenge

A nice challenge with a remarkably simple solution discussed at length here .

mrvn wrote: Thu Sep 08, 2022 10:30 pm One Covarex centrifuge with 2x productivity modules 3 surrounded by beacons produces an U-238 every ~8 seconds.

Well, there's 2 reasons that just ain't gonna hunt in my eyes.

A u-238 every 8 seconds is wholly insufficient for ammo production. That barely covers my average consumption and ammo demand is very surgy (surgish? surgilicious?). Demand spikes when you're on the offensive or when you're establishing new military outposts. Demand *really* spikes when you first build the stuff and you're trying to recycle all those red magazines into green mags and replace all the magazines on your perimiter. Also a pretty big spike when you get your hands on artillery and you're adding that to your perimeter. Both of those really big spikes happen before I even have beacons which leads me right into point 2.

I'm definitely not inclined to wait for beacons! I put up a running nuclear reactor while I'm still on blue science and I'm definitely not doing purple science before gold.

Gold has logistics, uranium ammo, portable fusion, power armor 2, military 4, a slew of damage upgrades and artillery.

The only thing I can think of that's gated behind purple that I may really want urgently is coal liquefaction in the event it's a very oil poor map. Beacons, blue belts, assembly 3s, and tier 3 mods are nice but just kinda endgame stuff.