All Master Class Blueprints are available on FactorioBin Overview and direct links to all Blueprints: nilaus.atlassian.net/l/cp/HBEUm524 (Pastebin links no longer work)
You kinda let me down here with nuclear blueprints Nilaus, I can understand steam storage versions are bad but some people want their powerplant to shut off when they're not using the power, I was expecting you to dig into the challenge of creating the circuit logic since there are so few tutorials on it.
I took one look at nuclear power and thought "Nope! Sticking with my 264 boilers, 528 steam engines, 14 pumps, aannnddd lots of drills." But after seeing this, nuclear is much less ominous to think about! I appreciate your video a lot :)
Same for me! Really the only thing that is annoying to set up are the drills, because (for me at least) my mining drills are very far away from my dedicated power generation area😂 but yeah its so much easier to set up than I originally thought
Sacrificing the ratio slightly to make everything more organized is definitely something I hadn't considered, but boy does it result in a nice design. Love it.
This was my favourite Factorio Masterclass yet. I'm not a fan of tutorials that end up feeling prescriptive, showing designs and saying "use this blueprint" rather than talking through the underlying theory and tricks being used, encouraging the viewer to try out their own ideas. I really liked how this time you deep dived into the numbers and reasoning, and provided good insight for experienced players and those following the wider community trends. In my main factory, I put off building a nuclear power plant even with a high effective Kovorax processing system*, because the effort to design my own power plant with all the adjacency bonuses was too daunting compared to the ease of just letting my bots place another solar farm. Hopefully this tutorial will help future players feel less challenged by the switch to nuclear! * When you cover Kovorax processing, please show how no circuits are required. There's very simple circuitless setups, and you really don't need any of the complexity of circuits. The one I use is simple, tileable, beaconed and doesn't need any circuit logic, and ultimately far far too effective.
Yeah, I have an artillery outpost really really far away from my base, and I train off uranium ammo, artillery shells, and obviously, steam to it and it works really well!
Although this master class series tackles some concepts that are mostly useful for new players, I can honestly say I learned a ton from this one. Thanks very much mate, keep em coming!
Hi Nilaus, Hope you are doing well! I'm aware this video is a little old. But I just started with Factorio and I'm at the stage where I'm setting up nuclear power. This guide and your blueprints are just GOLD. Like Jesus, this helped me out SO MUCH!! A very big thanks for your help and effort, you Mr are GOLD!!
Just here to confirm that I've just built a 10GW nuclear power plant, with almost exact ratios, that works at peak capacity indefinitly as long as there is fuel (it really does work, its not those ones you find on the internet that works for a few minutes before running out on water). I did it only cause I wanted a very big one, not because its highly needed. Your points are tottally valid.
You're right about the 2x2 going to a 2x3 reactor grid that it consumes just 50% more fuel, but your power output goes from 480MW to 800MW, almost doubling the power. So very much worth it to make such a design. I made a design which is almost a perfect mirror, it even includes roboports and has 1 belt going through the design which supplies the fuel and carries the spent fuel out as well without letting the input fuel passing through, it's stopped right after the reactors. And my design also lets you tile it perfectly in case you need more power. Both belts run parallel to the steam engines, one supplies the fresh fuel, the other gathers the spent fuel and they connect nicely. I may have to redesign it someday because my roboports aren't perfectly on-grid atm.
hello, here's my 2 cents : - while i'm 100% OK with dumping "perfect ratios" when they're not relevant, (so, most of the time :D) - in the 4 reactor design, i'm very surprised that you choose to use the maximum potential of offshore pumps which are so cheap and trash the excess nuclear reactor potential which is so expensive. I usually go like this : 4 reactors + 6 pumps -> 6*8 heat exchangers > 6*14 steam turbines. Too many pumps & steam turbines, but every bit of the reactors power is tapped (i get 480MW out of it instead of 464 MW) - doing maths makes people leave ? what ? that encouraged me to STAY 8) - i agree with most of the opinions of the second half... although nuclear power without kovarex relies on having enough space to store excess u-238, and that's more than 1 steel chest per hour PER REACTOR, even with prod modules... so should it be advised to beginners ?.. - nice video overall, well explained, well organized.
kirkmcdonald.github.io/calc.html#data=0-17-60&rate=h&min=4&k=off&items=used-up-uranium-fuel-cell:f:1 According to this, fueling 1 reactor without kovarex yields an excess of 221 U-238 per hour, or 21.7 hours to fill a steel chest. That's plenty of time to tech up to kovarex and set it up.
I should have read your comment before I built the reactor exactly as shown in the video--except that the water was a bit further away--and then spent a couple of hours trying to figure out why I couldn't get the full 464 MW of power. I noticed I was seeing 'Low Input Fluid' in several of my heat exchangers. It was easy enough to add a 5th offshore pump to make the design start showing the full 464 ME of power.
Came here only for the blueprint for Nuclear Reactors. Ended up with abundant amount of knowledge. Every single video you've done, you never disappoint
Overall I like it. - Totally agree with your opinion on steam storage, waste of effort. - Not sure I agree with your conclusion on 2x2 layout vs infinitely tile-able, but you make a good case. - The only thing I really disagree with is capping your design on offshore pump capacity. I would increase to 2 pumps per row (for a total of 8), or do 3 pumps per 2 rows (for a total of 6), so the reactors are your cap. I agree that nuclear fuel is cheap, but so are pumps. - I definitely agree that you can run nuclear power for quite a while before requiring kovarex, just stashing 238 in a dozen steel chests for a while. - 28:20 maybe some people, but when you said math my ears perked up! Loving this master class series!
I remember my first time playing factorio and i couldn´t (and didn´t want to) figure out the nuclear reactors, so i just stamped down a random blueprint from the internet. I think with that tutorial many new players can learn it fast, great tutorial! Wish you did this when I was new to factorio :D
i cant believe i just watched a 40 minute video about factorio nuclear power plants... and i gotta say, not a single minute was not needed great vid thanks for the help
One touch I like to add to my nuclear reactors is to have the fuel creation, fuel reprocessing, and the inserters for the reactors all on an independent power grid run on a few solar panels and accumulators. If you make a mistake that results in a complete blackout, it can be a nightmare to get things started again without that.
i usually tap off some heat from the end of the reactor rows, closer than any other heat exchanger and with an own water supply, plus backed up by a Boiler fed from a chest with some surplus burnables, and an Accumulator, both with a Global Alert if it ever gets discharged/used. This powers everything directly related to the reactor (Inserters, Combinator stuff, Pumps), and also the Uranium refinement Process. The Output Turbines are then on their own Grid(s), completely isolated from the reactor-internal Grid. (a few Turbines on the Edge might feed both Grids to allow for higher consumption on the internal Grid, but i think thats not necesary)
I would say that applies to coal/solid fuel power much more so. If you screw something up and lose power, kickstarting nuclear reactors isn’t that big of a deal you should be able to just clear out a clog and then manually fuel the reactors and it should get going again long enough for you to find the issue and correct it. I just keep a buffer of nuclear fuel and uranium if I need to do so. If your large coal/solid fuel generation facility blacks out (say you accidentally redirect a main fuel belt) getting that going is a pain because you need to feed coal to dozens of reactors long enough to get fuel produced and to the boilers, which is why I like to have a fuel supply on its own circuit that needs relatively few boilers to maintain that can easily be fed manually or with a single chest.
A few remarks: You are using assembling machine 3. Those are not unlocked at blue science, but at purple science. With only blue science, you can only use 2 modules per machine (8% prod instead of 16%), reducing the amount of fuel you gain from an ore patch. It doesn't change significantly your conclusion, nuclear power is dirt cheap in fuel and should absolutely be rushed in the mid game (as soon as you can afford the enormous oneshot setup cost and you need the enormous power generation). Productivity 3 is completely overkill in general, unless you already have a massive base. They cost a massive amount of resources for very little gain. By the time you can afford to productivity 3 your nuclear setup, you most likely have Kovarex. Using steam storage is mostly there to "future proof" your design (so you don't have to come back before a while) and to protect your grid from overloading because of laser turrets. Tanks and combinators cost next to nothing when compared to the enormous cost of the nuclear plant itself, so frankly there is little reason to not do it. If you wonder about UPS, DON'T DO NUCLEAR AT ALL, do solar. Solar cost nothing in UPS, that's why all the megabases are using solar. Once you have nuclear power up and running, you should upgrade your armor piercing bullets to uranium bullets. It only cost you U238 (the useless one) and uranium bullets does 3x the damage. They will shred through any bitters. You can do the same thing for your tank shells. Don't bother with nuclear missile however, those are next to impossible to build before Kovarex.
It's very cool to see that you actually don't need to do anything complex to support nuclear reactors, but part of the fun I get out of Factorio is optimizing every last thing down to the stuff that doesn't even need it
apropos the steam storage; I use steam storage and 20-50% extra turbines, to handle peak variable power load from laser turret arrays. My resting factory load is like 1.2gw but when the wall's getting pushed that can peak at nearly 2gw so that extra capacity just being there ready to go is v v nice, and the turbines/steam storage aren't terribly expensive especially compared to battery banks.
They are incredibly expensive in the Update calculation side of things though. Each pipe and steam engine needs to do their own calculations. Battery banks are the exact opposite in that no Matter how many you have, they will always need the same amount of updates. So if youre building really big or your PC sucks, use batteries and preferably solar over nuclear.
@@VoidplayLP oh yeah, that's a thing. I'm about to set up a Ryzen based system next month so I'm not super worried about it but yeah, my mega base lags a bit.
Thank you for another excellent master class series. I really enjoyed your math and motivations on why anything larger than 2x2 plants are irrelevant. Your arguments that a decent sized uranium patch will essentially provide unlimited power is very convincing. The modular design of these nuclear power plants are beautiful and I will be sure to use them in my next Factorio playthrough!
But the kovarex isn't for needing less miners, it's for allowing you to have that one patch last LONGER so you don't have to constantly run around and find new ore patches for your fuel.
Kovarex is for disposing of excess 238 automatically by turning it into 235. It's not about making patches last longer, it's about making damn near everything you dig up stop being useless dead weight.
@@williambarnes5023 You're both talking about the same thing. When almost 100% of the patch is U-235 because of kovarex, it lasts longer. You aren't mining lots and lots of ore just to use 0.3% of it, you can use almost all 100%. The U-238 that used to take up space, and that makes up 99.7% of what you dig up, can all be made useful. Everything you dig up stops being useless dead weight... and thus you get the maximum amount of usable U-235 from every patch. Having 100% of the result be useful absolutely means a patch lasts longer, since you're using all of it instead of the vast majority being dead weight.
I agree with your point but I'll set the numbers right: It's 99.3% (or 99.28% according to some sources), and slightly less if you module the centrifuge, so ~135:1, give or take 5. Kovarex needs a substantial working set (i.e. more than the stuff it actually converts) but the net conversion is 3 U238 to one U235. Not sure how a prod mod could change that, but anyway you should use them if you can afford the power and production speed hit. Anyway, first-hand U238 is converted to ~45 times as much U235 than the first-hand U235 you get, and Kovarex ain't random. Finally, second-hand U238 from spent fuel reprocessing can be turned into even more U235, but that doesn't increase the 45:1 figure for reactors - because a batch of fuel rods takes 10 U238 to make; you're merely reclaiming that from spent rods. (Not sure how a prod mod would change _THAT_ tho...) Anyway, I got here because the math looked off to me -- 1 _thousand_ ore per first-hand U235??? But that's right; the centrifuge takes a batch of 10 ore units per unit of U235. It's essentially more than just the centrifuge. It separates uranium from acid, tailings, and other impurities before separating the isotopes. So, one miner at 18 ore per minute makes 1.8 batches, and ~0.013 of those are U235. Which is enough to make 10 (or 11.6 with prod magic) times as many fuel rods, or ~0.15 of them. Each can run the reactor for 200 seconds, so you get ~30 seconds of reactor runtime out of a minute of mining operation; two would keep the reactor more or less supplied 24/7 while ore deposits last. Multiply that by 4 if you want to multiply the output of each reactor by 3, and you get the "8 mines" bottom line which Nilaus mentioned. (His treatment was a bit confusing because he didn't mention the ore shrinkage in the centrifuge, just that 0.7% (of the total output) is U235, and a bogus "18 per second" figure when he meant per minute. I hope my math is on point. I'm interpreting the "8 miners" figure as the requirement per plant of 4, rather than a figure per reactor block.)
Funny, I grabbed the biggest uranium patch and put 177 miners on it and then had 52 centrifuges going. Once I watched and built this, I realize that is way overkill and there is no chance of my power plant ever running out of fuel, even if I duplicate it a lot. Thanks for the awesome info as usual!
Another great design that is very efficient in space and power. My nuclear is blueprint of choice until now has been the KOS nuclear plant. It is modular, but not nearly this compact. Thanks for the fresh look and design!
Excellent design and it seems you came to lot of the same conclusions as I did when doing my nuclear setup (although I never did the miners per reactor calculation, that is kinda crazy). I also ended up with a 2x2 reactor setup (since you've got most of the neighbor bonus by then) and pulling water from one side without circuit controls or steam buffers. I sometimes add circuit control when I run these before unlocking kovarex but just adding a single storage tank is sufficient to monitor steam levels. Full steam storage is totally unnecessary. It is good to see such a useful design getting highlighted. Cheers!
argument for steam storage: unexpected power surges (or not looking at power consumtion) that enables you to put more steam turbines than the heat exchangers can usually handle thus allowing for a compact battery
This reactor design produces 480MW(th) out of which 464MW(e) is used to produce electric power. A4W reactors on US Navy Nimitz-Class Aircraft carriers produce up to 550MW(th) each (1100MW total), of which only about 100MW is used for electricity production (wikipedia), the rest of the capacity is used for ship's propulsion, plus some extra capacity for safety margin and sustaining the aircraft carrier with one of the two reactors down for maintenance or recovery from a casualty. This design uses reactors that, unlike A4W reactors, operate at 100% capacity at all times. When operating A4W reactors, we typically maintain power as low as practical for the ship's operating conditions. Let's just say with both reactors running and normal day-to-day operations when in a mission area, they're typically operating at less than 50% each. The reason for using both reactors when one could carry the load demanded of both is to have operational flexibility if one reactor needs to go offline for maintenance or in response to a casualty. Factorio reactors also use fuel enriched to... exactly 5% (1/20 parts U-235, 19/20 parts U-238) whereas the enrichment of navy nuclear reactors is... I forget exactly but it's legit well over 90%. It may be 98-99% even. This allows navy nuclear reactors to load enough power-production capability to last a ship for literally decades. Someone with a lot more dedication than me could probably calculate the mass of U-235 consumed in 200 seconds to produce... well... actually, the fact that factorio reactors have this neighbor bonus thing kind of throws off the calculations. Why does a separate independent reactor adjacent to another reactor cause the fuel to produce more usable energy?
Maybe the neighbor bonus is modelling a larger combined reactor instead of many smaller adjacent ones. In a larger reactor core, the otherwise wasted neutrons could be used.
I need to pull the blueprint, but I spent around 20 hours one time a year or so back trying to design the perfect nuclear power system. the end system used 6/8 reactors, double heat pipes, tanks for water and steam but not for stream storage because there was exactly 0 fluid dynamics checks because everything was moved by pumps. the water came in through the middle, between the turbines and the reactors, and across the bottom. if you flipped the blueprint 180*, you could stick another one on the bottom and get a slightly higher bonus, and if you put them side-by-side they could share heat to speed up the initial heating process, and would share a mutual water input stream. the pumping system could handle just less than 3x8 reactor setups horizontally before it would start to have problems getting enough water into the heat exchangers at the back. the last heat pipe held a constant 525 degrees when the system was fully hot and fully utilized, which is why I chose to stick to 6/8 reactor systems. can't get the heat out far enough to need any more than that. I think that's the earliest hard stopping block. when I say 6/8 reactor setups, I mean 8 to start untill all steam tanks are full, then take the bottom 2 away to leave 6 which maintain enough output to get 1.2gw steady, but built with enough turbines to do just about double that because that's what the system should be able to handle since it's duty cycle is only about 50%. the best and only argument for steam storage is the day/night cycle and solar panels. Free(ish) power, and you definitely shouldn't need all of your reactors at full steam during the day if you aren't relying solely on them for your whole base's power. why not bank the rest of the steam for later? it's only a few tanks. pump in pump out and you won't have to worry about fluid dynamics checks. sure it's a bit bigger but, like, is your map size limited or something? just tack on a tank field connected by turbines to burn off the overflow in times of very high usage. obviously I used bots to deliver the fuel. I mean, who would even consider trying to belt that stuff in and out. lol :P (I kid) getting water in was always the problem, but can be overcome with proper placement and delivery techniques (pumps mostly) All in all, that's a great guide though. well thought out and reasoned and the design is great for a 'bases first nuke field' setup. the math may work out to being able to do about 1 reactor to 8 miners, but I'd like to see that actually work in practice. what would the power consumption of 8 miners and (I think) 8 centrifuges with 8% productivity modules be compared to the power output of a reactor? what's the real power profit there? or is that just another really roundabout way of banking solar power when you think about it?
@Tommy Taffy has been a long time since I was a factorio master, but, when I designed this fluid dynamics were far more rudimentary than they are today. at the same time, it's likely that pumps bypass the standard fluid dynamics that seek to balance population and flow between connected pipes and fluid containers. pumps move fluid from one vessel directly to another with no mind to balancing or viscosity. if you chain pumps, or pump->vessel->pump, once filled, will flow at pump speed throughout the system (minus any fluid used or forked in previous steps). fullstop. the gain in flow rate and fluid availability in deep systems is not negligible. try it for yourself.. chain pipes, underground, tanks... then do pump/tank/pump and see how much faster the end tank in the pumped system fills over any other layout.... then to prove it to yourself, in the middle of the pump setup, switch 3 pumps for 2 tanks and put 3 in a row in the middle of your loop, or connect with pipes instead, and see how just 1 fluid dynamics calculation substantially slows the rate the end tank fills. I'm going to dig up that blueprint... I'm sure I still have a vanilla world from 3 years ago with it built somewhere.
I get it and I knew about the abundance of fision material since the first time I played the game (because I can read). I still don't like running my power plant at 1000°C.
Here is why I disagree with your perspective about steam storage: One: steam storage is eeeeeasy. Just slap some tanks down in the free space along the sides of your boilers. The design difficulty is zero, and compared to the cost of the rest of the parts, a dozen tanks is basically free. Two: peak capacity being higher than sustained capacity is useful for e.g. situations where you are transitioning from heavy solar or have lots of laser turrets or other situations where your power consumption needs might be "bursty". Three: reactors which are hot (>900C) EXPLODE if they are destroyed by damage, with the same effect as a nuke going off. Nothing like a few biters getting into your power plant, and then the entire thing literally blows up and you have to rebuild all of those extremely steel heavy structures on a factory with no power. Four: It adds resilience to brief outages. If I need to disconnect the water supply to re-route pipes, I can do that, and it will run for 10 minutes. Technically, storing heat directly in pipes is more space efficient (500MJ/tile compared to steam tanks at ~269MJ/tile), but just building more heat pipes in random places is ineffective because as you point out, heat transfer becomes a bottleneck. I disagree with your perspective about circuit-limited reactor designs too, because they are very simple to implement well. I homerolled a controller design which is 3x4 tiles, contains only 6 combinators, and it's both jam-resistant and idiot-proof. It's the difference between waiting until kovarex to spin up nuclear fuel for your trains, and starting it as soon as you can.
You make good points. I guess it depends on the situation: a big base with low UPS means not using tanks will make the game run faster, because it doesn’t deal with the fluid calculations involved, and a base on a limited-resources-run is way better off implementing efficiency measures. It ultimately comes down to preference and convenience.
Well, if bitters chew your nuclear reactor, you have something very, very wrong with your base defence anyway. I do agree that steam and circuit based nuclear power are simple to build, and they are scalable. You can easily build a 480 MW power plant (that's 4 reactors) even if you only need 60 MW right now (it's usually the power I need when I start looking into nuclear power). You won't use that much fuel, stockpiling for the future, and you won't need to deal power for a while.
@@giovannidueck9094 Megabase skip nuclear power entirely and only use solar. Solar have no UPS cost. Nuclear power can't beat that, with or without steam tanks.
Regarding point 3: With the addition of the spidertron, this has become a non-issue. Simply building the reactors on an island that is only accessible by spidertron makes them completely invulnerable.
Doesn't really matter once your Kovarex process is up and running, as you basically get unlimited nuclear fuel. Get a couple uranium patches being mined and processed and it's pretty much impossible to be short on it.
Great guide! It's really surprising how few miners and centrifuges are needed even without kovarex. I always start doing on-site uranium processing as soon as I unlock the research. I don't use modules, so I go by a 10 miners : 3 centrifuges ratio. I usually start with a 2 reactor build because its cheaper to build early on and satisfies my power needs for a long time, and just the 10 miners and 3 centrifuges is enough to support it indefinitely. The nice thing about having a tiny uranium processing build is that you can put efficiency modules in everything and never have to worry about biters. Sometimes I will use steam buffering for the 2-reactor setup because you only need a couple fluid tanks (maybe 10 or 20) and the wiring isn't too bad. For the 2x2 build though its definitely not worth it, especially after you get Kovarex.
Lemme put it this way, I LOVE this design. It’s aesthetically pleasing, works brilliantly and is easy to put down. Then a friend and I started playing with the Space Exploration mod, and this became INSANELY useful. That reactor blueprint is as vital and as much a useful a tool for us in that game as train stations are, we decided.
There is a good way to measure the actual power produced with an "electric-energy-interface" (you have to create it from console). When you set it to consume power, it has the lowest priority, same as an accumulator. You'll be able to see on the charts how much power is going into it. This is nice because it not only reveals flow issues but also includes all the pumps or whatever you use to keep the reactors running.
An interesting point is that even if you're committed to steam storage for your reactors, you probably only need to build your first reactor with steam storage. You'll be building a second reactor because you're about to exceed the capacity of the first one, so every other reactor you build can be running at full load without "wasting" power anyhow.
This is what i do, I rush nuclear to build a 2x2 reactor with removable steam storage, I use the steam storage function while my base is under 480MW and remove the tank section when i hit full load and paste down another 2x2 reactor when I need more power. I know wasting uranium isn't really a concern but i like the puzzle of it and like not wasting things.
I have two extremely efficient designs that "just work", a 2x2 and a 2x5, both of which fit nicely within Nilhaus' city block blueprints. They feature auto start/stop for power saving and peak power boost thanks to steam storage tanks, backup power for the pumps nearest to the water input (assuming you disconnect their wires), all without a single combinator or requiring any sort of tuning. The 2x2 has a max boost power of 500MW, and sustained 464MW of a possible 480MW, and the 2x5 has a max boost of 1.6GW, and 1.37GW sustained out of a possible 1.44GW, so they are highly efficient too. I tested them rigorously to ensure they do not suffer from any of the common pitfalls that Nilaus mentions in this video. You can of course remove the steam storage if you don't care about fuel efficiency or all those nifty features. They are available on the Factorio Forums here: forums.factorio.com/viewtopic.php?f=208&t=91639&p=521863
The point about larger reactor setups (2x3 and larger) is only about uranium-saving is very helpful! Now I can use my 2x2 designs peacefully without worrying that I missed something by not going to 2x3
Just got into factorio two weeks back and your videos have been such a great help! I really liked this video cause you explained your methodology. Keep the great work up!
Thanks Nilaus, you completely convinced me to not use steam storage. My little 600k uranium patch with a fully beaconed kovarex should last forever. :D
Hey Nilaus, this was a definite master class. Factorio graduate studies vs your 4 year first degree. Btw it is so cool when things like your uranium non-scarcity math appears on YT and I can remember being in the twitch stream when the calculation first happened ((: Bravo! EDIT: I’m SO used to seeing radial symmetry nuke plans designs in your playthroughs. My eyes almost can’t believe this version. That’s the real controversial opinion hehe
If you wanted to get that missing bit of water in your ratios you could use a storage tank for water for each set of heat exchangers and use two water pumps per storage tank to keep the water storage tanks filled. Gives you the missing water and would still be tileable.
One thing I like to do is add steam storage tanks and wire the system to feed fuel cells only when the steam storage runs low. This way, I use the fuel cells efficiently.
@@littlebuilder8132then just build 1x1 reactor (lone reactor) setups if you think uranium is so plentiful that you don't need to be efficient. I love Nilaus, but his argument here makes no sense. He is saying to build 2x2 for efficiency, but then don't build steam/circuits because uranium is plentiful so you don't need to care about efficiency. Can't have it both ways.
you really don't need to be efficient with uranium once you have an enrichment process researched and running. you lose on a lot of free power if you build just singular reactors @@beepbop6697
It is only tileable if we disregard water supply. If base needs several GW of power it gets real messy real quick. I prefer more spread thin design, it also allows for +3 adjacency even though it is not a big deal. Also, it would be excellent addition to show the design in action, to demonstrate that it is indeed producing at maximum capacity. Especially after what you have said about difficulty of designs with fluids.
I haven't ran into any lag issues concerning nuclear power plants yet, and because of this I have developed my nuclear power very well. I designed a nuclear reactor that is actually tileable, and as long as ample water is provided takes full use of neighbor bonus. It consists of a base 2x7 reactor that then can be made into a 2xHoweverManyYouWant in a row. Only caveat is it must be built on top of a lake filled with landfills, since it would be near impossible to supply water with pipes. In my solo world, I was able to build it before I launched my first rocket, and each tile generates I believe 2.2 GW, which is amazing since you don't have to worry about power and if you do use more than that you just slap another tile one it. I have also developed a very efficient kovarex refining method that involves using circuit network signals to regulate exact proportions of u235 put into the centrifuges. It does require logistic bots, but if they are available to you you can easily convert thousands of 238 to 235 to be used for fuel or, my favorite use, spidertron with nukes. Of course, using this would leave you with very little 238 for uranium ammo, but if you use lots of laser turrets like me its not much of an issue. If its requested, I can post my blueprints for either, but I'm sure there are already tons out there. Also, my reactor design requires lotsss of landfill (it has to be built ON, literally ON TOP OF a lake) and isn't going to serve as a beginner reacter lol.
Space Age comes out next week and I figured I should probably learn how to go about nuclear power and the purple and yellow science packs. Even now, 4 years after this video was made, it’s still very useful - it’s hard to figure out the ratios. (Hopefully Space Age’s production rate display helps with this and Fusion Power)
Space Age just made the argument of no kovarex or steam storage stronger, as there are now miners with Resource Drain stat , which is a new seperate bonus from Productivity, but it goes down to 16% drain on regular miners (so 84% of the time the ore patch does not go down), and starts at 50% for Big Mining Drill, up to 8% (so up to 92% of the time, it makes ore not consume the area) Quality on Reactors/Heat Exchangers and Turbines also are quite interesting but I see it mostly a way to reduce space and also further reduces fuel usage, but needs calculations if mixing different quality levels. Also you can now read the reactor temperature via the circuit network, so you can set it to insert (or empty the fuel cell) only above/below certain temperature, as heat pipes will continue to work as long as they are above 500c and temperature can go to double that. (so if it's a reactor setup , but only using partially, the temperature gets used up slower, so fuel gets saved by cooling down to let's say 600, and since they also added a way to read fuel inside reactor now, it never inserts more then 1 if you set it so. Also adding more pumps exchangers and turbines will be interesting when quality comes into play (changing ratios if quality is mixed) the but the new fluid system makes pipe length irrelevant (up to 320 length) as the length of pipe is one giant storage tank now. Kovarex is still helpful for other things, as nuclear missiles cost 100 U235 now
i have a 2x80 power plant for almost 20Gw. Tricky part is finding a perfect lake to build on it Thank for the fluid info, I will downsize a lot of storage.
I managed to build a 2 GW reactor (14 reactor cores) and got it working flawlessly. No steam problems, no heat problems, no water problems, ran at full capacity. I figured the efficiency would be nice, especially in the late game, but any more than 14 reactor cores and I start having issues that I can't resolve. The problem with a 2 GW reactor, as mentioned in the video, is that it's absolutely huge. I had like 5 of the damn things, best place I could find only fit 2 of them, the other 3 were just placed wherever I could find space. I think smaller builds probably would be better, despite being less efficient. I'm going to try building a compact 1 GW design for my next run.
8 reactors in a 3x3 grid, with one of the sides removed is actually the most efficient in terms of the neighbor bonus. Use long handed inserters to put the fuel in, and you're good to go.
I'm not sure if that's what the nuclear fuel recipe has always been but I swear I remember one of the main reasons you needed Kovarex was because of the sheer amount of U-238 that you'd have to store which was why so many people wanted an incinerator of some sort for the game.
Ah, I always wondered why he invented that.. It would make sense exactly in the situation that you describe. And it would also explain why we always have way too much uranium patches by default.
I don't have the funds to support, and thereby show my approval at this time. But I am truly grateful that I can benefit from your knowledge, effort, and passion like this. I hope that my gratitude counts as a small token. (But I also hope that you are getting worthy financial support as well). It's like the feeling when you see a street performer who is breathtaking, but you have no cash on hand. You feel almost like you've taken from them as you walk away, without having given back.
Thank you. Don't worry about it :) I provide this content for free and hope that a small percentage want to and is able to support. I do not want any paywalled content, so just enjoy it :)
I saw a Reddit thread where someone showed off their 2x5 build. They showed screenshots that demonstrated it was indeed running at peak efficiency, and it was absolutely beautiful, but my gods, I cannot imagine how many hours they spent on it. Unless you intend to just copy that person's blueprint, I would highly suggest taking Nilaus's advice here.
or use water well pumps (offshore pumps buried under landfill), provided you have access to a big enough lake, it solves the trilemna of heat/steam/water connection : just feed the heat exchangers lines directly with water well pumps and the rest will get far easier to deal with.
DL the badass BPs if you don't feel like doing it yourself. There's a 2.6 G All circuited up and it works perfect. So yeah, figure out your own designs but once you go big don't bother, it's already been figured for you, unless you just like that kind of challenge. Very instructive video for me. Thanks for making the video.
Steam storage and overbuilding turbines is good for dealing with surge demands (like a seriously oversized laser defense grid) Larger nuclear reactor setups save space overall. Start mining Uranium right away, save enough to start enrichment(not much honestly), then build your reactor fuel supply. When enrichment comes you can start right away and hit those milestones for nuclear weapons and nuclear fuel for your trains.
My entire factory consumed less power than a single reactor outputs (b/c solar) but I wanted better efficiency so built a 2x2 reactor anyway. No way in hell am I wasting all that fuel running full time. Definitely worth the circuit effort... though I needed to make a T-flop train circuit to keep accumulators charging on only solar. That part was a mess lol
"if you are not an absolute master in designing these, don't do it, its not worth it" Its a designing game, you design something and work out to increase its efficiency.
Well, I think it depends on what your goals are. Some people like to make the larger setups just to do it and because they enjoy the design process; others are more practically oriented, so beyond a certain point if it doesn't actually further benefit the functionality of the factory it isn't worth doing.
@@strategicsage7694 If you do not understand how a setup/blueprint design works; you shouldn't use it. Thats almost all the fun of this game. I don't know a single person who enjoys factorio just by copy pasting some blueprints off the internet and launching a rocket.
@@kitsukeita Where do you draw these people you know who play factorio from? That's certain not what I observe. I see a lot of channels where the biggest request is for the blueprint downloads. A number of people specifically say it's because they can't be bothered to make their own designs and/or just want to get past whatever aspect is causing them trouble. People play Factorio for a lot of different reasons, but there's definitely a sizable group of people who are more interested in the solution (blueprint) than the understanding of it or the design process.
@@strategicsage7694 People might ask for blueprints for various reasons, maybe they want to assimilate it in their own design or they want to further improve it or they just like it more than their own design; but if you do not understand what the blueprint is supposed to be doing & how, i see no point in playing; might as well copy those small starter base prints from the internet(meant for advancing faster in modded games), hand craft the components/buildings, launch the rocket and call it a day. My data of people comes from my steam friends/factorio steam groups/Bobs,Py + some other mod discord channels.
Who was hoping the thing would blow up when Nilaus said he had never seen it before :) and you should add accumulators as well to store any excess energy since it was built to run constantly in this form. That would make it more viable and unlike steam storage this would allow it to run without losing any energy overflow.
I use steam storage with additional turbines as a means to store energy for turret spikes. I also setup an alert system so that once it is below a certain point(power use exceeds actual power production) I can go add another module to the power plant.
you can also use the steam storage to prevent over heating of your reactors : nuclear plants heat up to 1000°C no matter the consumption, any excess heat is wasted. If you setup things so that you only empty your reactors when steam storage goes bellow some level, and you only fill back when emptying, you will minimize fuel waste.
I happen to be a very inefficient player, so my factory backs up a lot and fluctuates heavily in consumption. Steam storage always felt right to me as they are essentially “better accumulators” but the performance argument makes sense. Exchangers produce slightly more fluid than the turbines can use which ensures the pipes are always filled. I tend to just go for “max efficiency/perfect rate matching “ when the better idea is to just overbuild everything as nothing is really wasted in this game
I wish I had seen this video a month ago. I have a space exploration megabase krastorio 2 game going right now. On my main planet I ran out of all nearby uranium sources and had to start importing it from other planets because I actually did my Kovarex too much/fast/whatever and completely drained all of my bad uranium. I had this huge panic when I saw my nuclear fuel generation had hit zero and it wouldn't restart because I didn't have all the needed materials. Thankfully, I had stockpiled centuries worth of fuel to keep the base defense lasers going long enough to start getting cargo rockets of uranium sent back home. TL;DR - Thanks for the math at the end, super cool stuff I wish I had known forever ago.
For people watching this after 2.0, making these fuel efficient became trivial now that circuits from reactors emit a fuel signal and a temperate signal. A very basic circuit can make 1 uranium fuel last many orders of magnitude longer.
@@Klorel123 It's not letting me post a link for some reason. So here's a summary of what I did. You can DM me if you want a link to the blueprint. It's very straightforward. You just make each reactor output temperature and fuel signals. Then, for each reactor, you just have one decider check the temperature signal and output 1 on H if the temperature falls below some threshold (I chose 650°C); and you have another decider check the fuel signal and output 1 on Uranium Fuel Cell if the fuel from the reactor falls below 1. Then you connect both of those deciders to the input of an arithmetic controller and have that output H AND Uranium Fuel Cell as R. Attach the output of the arithmetic controller to the inserter that puts fuel in the reactor and check Enable/disable and set the condition to R > 0 and override the stack size to 1. Do this for all four reactors separately. This makes it so that the inserter will only load 1 fuel cell into its reactor when the temperature is below 650°C AND there's no burning fuel in the reactor.
@@aloric711 I'm just trying to learn logic in the game, can this not be done with one decider using temp AND fuel cell conditions to output 1, then the output of decider to the inserter? so need need for the arithmetic?
Using those numbers, we should only need around 2 minuers per reactor. Either the ratios you presented aren't correct, or else I made a mistake in my calculations somewhere. Can anyone help with this? So we have: In the miners: Uranium ore per miner per minute = 18 Uranium ore per miner per second = 18 / 60 = 0.3 In the enrichment process: Amt. of Uranium Ore per unit of Uranium 235 = 1323 In the assembling machine: Amt. of Uranium 235 per unit of Uranium fuel cell = 10/116 ~= 0.08 In the reactor: Amt. of fuel cells per reactor per second = 1 / 200 = 0.005 Putting them together: # of miners to constantly power 1 reactor = (1/Uranium ore per miner per second) * Uranium ore per Uranium 235 * Uranium 235 per Fuel Cell * Fuel cells per reactor per second = (1/0.3) * 1323 * 0.08 * 0.005 = 1.764 This means we need less than 2 miners per reactor, as opposed to the ~5.6 miners mentioned in the video.
I think he said it wrong, it's 7.6 miners for the block of 4 reactors. Nilaus often misspeaks but then he is not a native English speaker. (Although it really irks me that he calls "Sulphuric acid" "Sulpher" ALL THE TIME! ) Doesn't detract from the amazing builds and tutorials he produces
Hmm, this is interesting. I didn't know *just* how abundant uranium was, though I was aware it was ridiculously abundant. I enjoy designing and have made some ridiculously large power plants. I've done a 2x16 model and I think it's the best way to learn about water and flow mechanics. The amount of pumps just to increase flow for the water I had to use was staggering and the steam required a lot of attention too. I've ended up with hundreds of thousands of Uranium, both shiny and dark, and I guess I never really thought about it beyond knowing that between steam storage as batteries and efficiency from the amount of reactors I was using that I was cutting down on my uranium consumption significantly. I'll probably keep doing that. Definitely once I get to the point where I need more than 4 reactors, the design problems involved with nuclear is very entertaining :D.
The reason I use the refining process is to consume the 235. My biggest problem with nuclear power is the lack of ability to delete items. In my first time using nuclear power I set up 20 steel crates to hold all the 235 and once they filled up my centrifuges stopped pulling out 238.
Reminder to disable snap to grid on the reactor builds. They don't play nice if your blocks are aligned to the rails, but not to the very specific grid used to make it.
yiu might not need koverax but if you don't have have it you have to store U238 somwhere. Also if you don't use koverax to feed your reactors it will take even longer to get the U235 you need to start koverax since now your also using up some of it to run your reactora
Your best yet, Nilaus! I especially appreciate the theoretical background. I see your point regarding the calculations, but frankly I also see it as a challenge to do it better... :)
good design but inefficient. You only put in enough 235 to heat the exchanges. then take the 235 back out. Storing the Heated Steam is the most Efficient Method Period. All you have to do is create a simple circuit to register the 500/750/900c or what ever your mod is. Once the Exchangers hit target temp filter back out the 235, when target reaches below like 30% of max temp filter back in.
Thanks Nilaus, if I ever come back to factorio I will know for the future on how to do a nuclear set up. My first attempt at it worked but it wasn't perfect and for some reason I NEEDED steam storage for the damn thing to be stable production. Next time I'll use your design. Love your videos and designs bro!
I know this is an old video, but I still kinda want to discuss some things if possible. First of all while I do agree with many of your arguments regarding steam storage and big reactors, but they are based on the question of whether you need to save fuel or not. So I feel a bit dissatisfied about the argument that you never need to build bigger than this one based solely on the fact that fuel is cheap. 2x1 would make for an even simpler build and probably even easier to tile - it's not more complex than a typical coal burner setup, so why not use it? I think saying that 2x2 is all you need needs a little elaboration, on why not smaller. Personally, I think the question regarding the size of the reactor is about what you feel comfortable building yourself. On my last save I personally went for 2x3 reactor supplied by bots with nice symmetrical design, meant to be stamped all over the lakes. 2x3 is not much more complicated than 2x3 - requiring 80 heat exchangers and 138 turbines, and it outputs a nice round number of 800 MW Next I do kinda disagree with you design process of limiting your reactor output by water pumps, like other comments have mentioned. Water pumps are really not that expensive, and sometimes building more of them just produces a simpler design with less piping involved, especially if we are doing landfill design that's meant to go over lakes - you can just set up a pump for each row of heat exchangers, and leave it at that. So I personally overbuild my water pumps and turbines a bit for the sake of simplicity and symmetry - like for example my 2x3 design actually has 144 turbines instead of 138, because it consists of 8 separate sections with 18 turbines each, and I want each being able to give maximum output. I can connect them together and remove some of the turbines, but honestly I don't see the need for that and the current version is cleaner and aesthetically pleasing. The setup has also been tested, and it does produce exactly 800 MW at maximum capacity like it should. There is of course a consideration that you usually want to have more power available than your base consumes, so having your design limited by water pumps is probably pretty much irrelevant in the long run, I guess our difference here is that I prefer to round up and you opted to round down in this case.
Agreed. If the argument is that uranium is so cheap/plentiful then why not just build 1x1 (lone reactor) setups? Anything larger is for the purpose of getting the neighbor bonus which *saves fuel* . So on one hand he says you need 2x2 for the efficiency, and then on the other hand limiting the water (and no steam/circuit) because uranium is cheap and plentiful so just ignore the inefficiencies. Can't have it both ways.
I can think of a reason to use steam storage on your first pre-kovarex reactor and that is to save U235 to get kovarex started ASAP. The less you use on fuel cells, the sooner you can have effectively unlimited U235 to make trains go fast and nuke every forest in your way, which is the proper use of nuclear power.
Thank you so much man for this tutorial, I was having so much problems with my last nuclear setup and I was running from here to there trying solutions to patch them but when i think i had fixed one then another came up. This is a like and a new subcription
You rock N! This is what i needed to reacquaint myself with nuclear. stopped playing for awhile after 0.12 and wanted to refresh after 1.0 and all of this free time. you are excellent at explaining.. gotta say the statement '1 fuel to 200 reactors' is a nonsense statement but it sure is attention grabbing :) I've noticed in the past that there isn't any point in steam storage - except the challenge *grin*
For others, here's another way to break down the math behind "uranium-235 is more abundant than you think" is in terms of time. There's just under a 1% chance of getting a uranium-238 each time uranium is refined, which can also be written as a 1/100 chance (to be more specific, a 7/1000 chance). Uranium refining takes 12 seconds. If you had 10 centrifuges all getting a steady supply of uranium ore, they would refine 100 uranium in two minutes; if you scale back to a more conservative 5 centrifuges, they'd refine 100 uranium in four minutes. That means on average, with five centrifuges, you'll be getting a piece of piece of uranium-238 every four minutes, five minutes to lowball it if you're feeling unlucky. A single piece of uranium-238 produces 10 nuclear fuel cells, which in total will run a reactor for 2000 seconds, which is 33 minutes. So, once that first piece of uranium-238 pops out, it's going to provide you more than 30 minutes of power, and you want to get another one to make more fuel cells before time is up. But within that 30 minutes, you'll have around 6-7 "almost a sure thing" chances. And this is all without modules to increase speed or productivity, and on just five centrifuges. The odds are ridiculously in your favor that you'll get at least one uranium-238 in time, likely more than that, and of course as you keep playing the stockpile safety net will keep growing. The catch then is just making sure you've set up your output lines in such a way to get the 238 to the assembly machine making the fuel cells in a prompt manner, but that's no biggie.
My first reactor is always a single reactor using steam storage. Then my second is a 2x1 or 2x2 depending how my supplies are. Then I deconstruct all of my reactors and build a 2x3. I'll try just sticking with 2x2 from now on. Thank you.
Really great with the calculations of miners needed, and as a result of that, there is no need for more than 2*2 designs. Hope there will come a blueprint for Kovarex at a point too.
To be fair, ive been watching for ages and actually took notice to check it i was subscribed and i wasnt. I swear i subscribed a while ago, sorry Sir! Love the content it really is absolute mastery
Something one may find interesting is in real life the contetration of u235 is about.7% and the amount of u235 in reactor grade fuel is 5% the amount of u235 in a fuel rod in Factorio is 1 U235 and 19 u238 which is 1in20 exactly 5% so the fuel is actually reactor grade also sulfuric acid is actually used in uranium refining
That Maths just blew my mind.... How can it be so low?!?! :D How much excess of the "Waste Uranium" does that "8" Miner process produce per minute after producing Fuel? If I am not converting it to Ammo, I would possibly need huge storage values until Kovarex Refinement comes online. I might try the maths on that myself.
Ok I'm new to Factorio but been playing Satisfactory for a while now and there is some mechanics or way it work that you probably assume everyone knows that new player don't. Things you don't explain here is : 1. How many or how much sulfure do you need to mine the ore. 2. The build leading up to the nuclear fuel. How much constructor do I need for that build to work ? 3. What do you do with the waste product ? Uranium 238 and the fuel waste ? 4. What does Kovarex do ?
1) One sulfuric acid (wish he would stop calling it sulfur, which is a different ingredient) per mined ore 2) Not sure what you mean here. It takes blue science to research nukes, at that point you will have all the necessary assemblers. (mark 2 assemblers and oil processing, specifically) 3) U-238 can be used for very powerful ammo and for kovarex (see 4). The fuel waste can be reprocessed back to make more U-238. 4) Kovarex is a way to convert U-238 into U-235. It takes 40 U-235 and 5 U-238 and turns it into 41 U-235 and 2 U-238.
I went with a circuit and steam storage design because otherwise of dozens nuclear power plants would have been sitting at 1000 c degree all the time (maximum capacity) and still consume fuels. All other energy demands are fulfilled by solar arrays during the day. Would you rather have fuel cells be consumed all the time? Or occasionally be turned off? It is kind of like having regular boiler consume coal/tree/etc... all the time. Lot of wastage from not utilizing all of the boiler output (1.8 MW) and miniscule power to move items.
Going large scale late with multiple reactors I can see how not using steam storage actually might end up beneficial. But for your first reactor using steam storage and a simply inserter circuit costs you basically nothing but saves a lot on actively consuming infrastructure you need to support your reactor, especially since when you migrate to nuclear power or the next reactor level you will most likely not yet fully utilize the power available to you and/or have some solar power in place. Fully agree on Kovarex however. That's something you can put off until you want to fuel trains with nuclear fuel.
Thanks again for some more great content. Really appreciate your hard work and clarity on these guides. So helpful. Can I ask a question, though? I don't understand how you worked out that centrifuges require an average of 1.3k ore to be mined to produce only 1 piece of U235. My calculations work it out at 132.3, so that's makes processing of ore into U235 10 times more productive than you say it is (30:05). This being the case, it actually make your point that you don't need to bother with steam capture or kovarex 10x more potent! What am I not understanding? Am I missing something? Thanks! --- Here's my math, just in case. U235 mining yield @ 0.7% = 0.007; 8% production bonus on centrifuge = 1.08 centrifuge yield; the combined yield = mining yield x centrifuge yield; => combined yield = 0.007 x 1.08 = 0.00756 ∴ Combined avg. yield = 0.00756 Calculate the inverse to switch unit to U235 rather than ore; => 1 / 0.00756 ≅ 132.3 : 1 ∴ Avg. ore needed to produce one U235 is approximately 132
the uranium centrifuge recipe takes *ten* ore to produce *one* bit of uranium. it takes 142.85 (1/0.007) cycles of the recipe to generate one unit of u-235 on average, so 142.85*10 ore, less the gains from productivity, which brings it down to around 1.3k ore.
All Master Class Blueprints are available on FactorioBin
Overview and direct links to all Blueprints: nilaus.atlassian.net/l/cp/HBEUm524
(Pastebin links no longer work)
You kinda let me down here with nuclear blueprints Nilaus, I can understand steam storage versions are bad but some people want their powerplant to shut off when they're not using the power, I was expecting you to dig into the challenge of creating the circuit logic since there are so few tutorials on it.
I took one look at nuclear power and thought "Nope! Sticking with my 264 boilers, 528 steam engines, 14 pumps, aannnddd lots of drills." But after seeing this, nuclear is much less ominous to think about! I appreciate your video a lot :)
This. I never thought it'd be this easy. Everything that goes around with all side-techs and big size builds just distracts from this easy build
@@dragoneyr1632 The The
Same for me! Really the only thing that is annoying to set up are the drills, because (for me at least) my mining drills are very far away from my dedicated power generation area😂 but yeah its so much easier to set up than I originally thought
Sacrificing the ratio slightly to make everything more organized is definitely something I hadn't considered, but boy does it result in a nice design. Love it.
This was my favourite Factorio Masterclass yet. I'm not a fan of tutorials that end up feeling prescriptive, showing designs and saying "use this blueprint" rather than talking through the underlying theory and tricks being used, encouraging the viewer to try out their own ideas. I really liked how this time you deep dived into the numbers and reasoning, and provided good insight for experienced players and those following the wider community trends. In my main factory, I put off building a nuclear power plant even with a high effective Kovorax processing system*, because the effort to design my own power plant with all the adjacency bonuses was too daunting compared to the ease of just letting my bots place another solar farm. Hopefully this tutorial will help future players feel less challenged by the switch to nuclear!
* When you cover Kovorax processing, please show how no circuits are required. There's very simple circuitless setups, and you really don't need any of the complexity of circuits. The one I use is simple, tileable, beaconed and doesn't need any circuit logic, and ultimately far far too effective.
My sentiments exactly. Thank you!
Well said, I concur. Any chance you could share the setup you mention?
Yes!
Hey, could u pls share ur own system's blueprint then it may come as passion for us?
Wait, why do you need beacons for nuclear power?
I make a setup where I train off steam to different parts of my factory as power. Why? Because it is extremely satisfying.
Steampunk Factorio? 😃
Yeah, I have an artillery outpost really really far away from my base, and I train off uranium ammo, artillery shells, and obviously, steam to it and it works really well!
@@bobi_lopataruwhy not just run big power poles?
@@dylanb2990 he doesn't have to run power poles just steam with the train track, I can understand that it does sound really satisfying
Dude's out here carrying buckets of steam around 😂
Although this master class series tackles some concepts that are mostly useful for new players, I can honestly say I learned a ton from this one. Thanks very much mate, keep em coming!
Hi Nilaus,
Hope you are doing well!
I'm aware this video is a little old.
But I just started with Factorio and I'm at the stage where I'm setting up nuclear power.
This guide and your blueprints are just GOLD.
Like Jesus, this helped me out SO MUCH!!
A very big thanks for your help and effort, you Mr are GOLD!!
Just here to confirm that I've just built a 10GW nuclear power plant, with almost exact ratios, that works at peak capacity indefinitly as long as there is fuel (it really does work, its not those ones you find on the internet that works for a few minutes before running out on water). I did it only cause I wanted a very big one, not because its highly needed. Your points are tottally valid.
how many reactors are there? or the ratios in general
You're right about the 2x2 going to a 2x3 reactor grid that it consumes just 50% more fuel, but your power output goes from 480MW to 800MW, almost doubling the power. So very much worth it to make such a design. I made a design which is almost a perfect mirror, it even includes roboports and has 1 belt going through the design which supplies the fuel and carries the spent fuel out as well without letting the input fuel passing through, it's stopped right after the reactors. And my design also lets you tile it perfectly in case you need more power. Both belts run parallel to the steam engines, one supplies the fresh fuel, the other gathers the spent fuel and they connect nicely. I may have to redesign it someday because my roboports aren't perfectly on-grid atm.
hello, here's my 2 cents :
- while i'm 100% OK with dumping "perfect ratios" when they're not relevant, (so, most of the time :D)
- in the 4 reactor design, i'm very surprised that you choose to use the maximum potential of offshore pumps which are so cheap and trash the excess nuclear reactor potential which is so expensive. I usually go like this : 4 reactors + 6 pumps -> 6*8 heat exchangers > 6*14 steam turbines. Too many pumps & steam turbines, but every bit of the reactors power is tapped (i get 480MW out of it instead of 464 MW)
- doing maths makes people leave ? what ? that encouraged me to STAY 8)
- i agree with most of the opinions of the second half... although nuclear power without kovarex relies on having enough space to store excess u-238, and that's more than 1 steel chest per hour PER REACTOR, even with prod modules... so should it be advised to beginners ?..
- nice video overall, well explained, well organized.
Oh, c'mon, after getting to blue science even a "beginner" should be able to stash the trash in a couple of boxes without hand holding, no?
kirkmcdonald.github.io/calc.html#data=0-17-60&rate=h&min=4&k=off&items=used-up-uranium-fuel-cell:f:1
According to this, fueling 1 reactor without kovarex yields an excess of 221 U-238 per hour, or 21.7 hours to fill a steel chest. That's plenty of time to tech up to kovarex and set it up.
@@FDGuerin You're right. I did the maths myself last time and I don't know where I made the mistake.
I should have read your comment before I built the reactor exactly as shown in the video--except that the water was a bit further away--and then spent a couple of hours trying to figure out why I couldn't get the full 464 MW of power. I noticed I was seeing 'Low Input Fluid' in several of my heat exchangers. It was easy enough to add a 5th offshore pump to make the design start showing the full 464 ME of power.
" 1 steel chest per hour PER REACTOR ", nothing a machine gun doesn't solve :)
Here I was setting up my first nuclear reactor, scratching my head, when lo and behold! Just what I needed.
Came here only for the blueprint for Nuclear Reactors. Ended up with abundant amount of knowledge. Every single video you've done, you never disappoint
i started playing a week ago. I like the setup, and it adds up compared to my calculations
Overall I like it.
- Totally agree with your opinion on steam storage, waste of effort.
- Not sure I agree with your conclusion on 2x2 layout vs infinitely tile-able, but you make a good case.
- The only thing I really disagree with is capping your design on offshore pump capacity. I would increase to 2 pumps per row (for a total of 8), or do 3 pumps per 2 rows (for a total of 6), so the reactors are your cap. I agree that nuclear fuel is cheap, but so are pumps.
- I definitely agree that you can run nuclear power for quite a while before requiring kovarex, just stashing 238 in a dozen steel chests for a while.
- 28:20 maybe some people, but when you said math my ears perked up!
Loving this master class series!
I remember my first time playing factorio and i couldn´t (and didn´t want to) figure out the nuclear reactors, so i just stamped down a random blueprint from the internet. I think with that tutorial many new players can learn it fast, great tutorial! Wish you did this when I was new to factorio :D
I can’t stress how happy I am you went into the math. I love having everything make sense and having a reason. Thank you so much!!!!
i cant believe i just watched a 40 minute video about factorio nuclear power plants... and i gotta say, not a single minute was not needed
great vid thanks for the help
One touch I like to add to my nuclear reactors is to have the fuel creation, fuel reprocessing, and the inserters for the reactors all on an independent power grid run on a few solar panels and accumulators. If you make a mistake that results in a complete blackout, it can be a nightmare to get things started again without that.
i usually tap off some heat from the end of the reactor rows, closer than any other heat exchanger and with an own water supply, plus backed up by a Boiler fed from a chest with some surplus burnables, and an Accumulator, both with a Global Alert if it ever gets discharged/used. This powers everything directly related to the reactor (Inserters, Combinator stuff, Pumps), and also the Uranium refinement Process. The Output Turbines are then on their own Grid(s), completely isolated from the reactor-internal Grid. (a few Turbines on the Edge might feed both Grids to allow for higher consumption on the internal Grid, but i think thats not necesary)
Unlike coal power, you are extremely unlikely to brown out with nuclear power.
I would say that applies to coal/solid fuel power much more so. If you screw something up and lose power, kickstarting nuclear reactors isn’t that big of a deal you should be able to just clear out a clog and then manually fuel the reactors and it should get going again long enough for you to find the issue and correct it. I just keep a buffer of nuclear fuel and uranium if I need to do so.
If your large coal/solid fuel generation facility blacks out (say you accidentally redirect a main fuel belt) getting that going is a pain because you need to feed coal to dozens of reactors long enough to get fuel produced and to the boilers, which is why I like to have a fuel supply on its own circuit that needs relatively few boilers to maintain that can easily be fed manually or with a single chest.
A few remarks:
You are using assembling machine 3. Those are not unlocked at blue science, but at purple science. With only blue science, you can only use 2 modules per machine (8% prod instead of 16%), reducing the amount of fuel you gain from an ore patch. It doesn't change significantly your conclusion, nuclear power is dirt cheap in fuel and should absolutely be rushed in the mid game (as soon as you can afford the enormous oneshot setup cost and you need the enormous power generation).
Productivity 3 is completely overkill in general, unless you already have a massive base. They cost a massive amount of resources for very little gain. By the time you can afford to productivity 3 your nuclear setup, you most likely have Kovarex.
Using steam storage is mostly there to "future proof" your design (so you don't have to come back before a while) and to protect your grid from overloading because of laser turrets. Tanks and combinators cost next to nothing when compared to the enormous cost of the nuclear plant itself, so frankly there is little reason to not do it.
If you wonder about UPS, DON'T DO NUCLEAR AT ALL, do solar. Solar cost nothing in UPS, that's why all the megabases are using solar.
Once you have nuclear power up and running, you should upgrade your armor piercing bullets to uranium bullets. It only cost you U238 (the useless one) and uranium bullets does 3x the damage. They will shred through any bitters. You can do the same thing for your tank shells. Don't bother with nuclear missile however, those are next to impossible to build before Kovarex.
The whole time I'm watching this, I'm also watching my Kovarex run, just making a _little_ more 235 each time. It's... calming, I guess, lol.
It's very cool to see that you actually don't need to do anything complex to support nuclear reactors, but part of the fun I get out of Factorio is optimizing every last thing down to the stuff that doesn't even need it
apropos the steam storage;
I use steam storage and 20-50% extra turbines, to handle peak variable power load from laser turret arrays.
My resting factory load is like 1.2gw but when the wall's getting pushed that can peak at nearly 2gw so that extra capacity just being there ready to go is v v nice, and the turbines/steam storage aren't terribly expensive especially compared to battery banks.
They are incredibly expensive in the Update calculation side of things though.
Each pipe and steam engine needs to do their own calculations.
Battery banks are the exact opposite in that no Matter how many you have, they will always need the same amount of updates.
So if youre building really big or your PC sucks, use batteries and preferably solar over nuclear.
@@VoidplayLP oh yeah, that's a thing. I'm about to set up a Ryzen based system next month so I'm not super worried about it but yeah, my mega base lags a bit.
@@VoidplayLP If UPS is a concern, you shouldn't be using nuclear power at all. You should be using solar. Solar cost 0 UPS.
@@Alphenis thats...what i Said.
@@VoidplayLP advocating solar for any purpose other than remote radar outposts is stupid simply due to the footprint needed
Thank you for another excellent master class series. I really enjoyed your math and motivations on why anything larger than 2x2 plants are irrelevant.
Your arguments that a decent sized uranium patch will essentially provide unlimited power is very convincing.
The modular design of these nuclear power plants are beautiful and I will be sure to use them in my next Factorio playthrough!
But the kovarex isn't for needing less miners, it's for allowing you to have that one patch last LONGER so you don't have to constantly run around and find new ore patches for your fuel.
Kovarex is for disposing of excess 238 automatically by turning it into 235. It's not about making patches last longer, it's about making damn near everything you dig up stop being useless dead weight.
@@williambarnes5023 You're both talking about the same thing. When almost 100% of the patch is U-235 because of kovarex, it lasts longer. You aren't mining lots and lots of ore just to use 0.3% of it, you can use almost all 100%. The U-238 that used to take up space, and that makes up 99.7% of what you dig up, can all be made useful. Everything you dig up stops being useless dead weight... and thus you get the maximum amount of usable U-235 from every patch. Having 100% of the result be useful absolutely means a patch lasts longer, since you're using all of it instead of the vast majority being dead weight.
I agree with your point but I'll set the numbers right:
It's 99.3% (or 99.28% according to some sources), and slightly less if you module the centrifuge, so ~135:1, give or take 5. Kovarex needs a substantial working set (i.e. more than the stuff it actually converts) but the net conversion is 3 U238 to one U235. Not sure how a prod mod could change that, but anyway you should use them if you can afford the power and production speed hit. Anyway, first-hand U238 is converted to ~45 times as much U235 than the first-hand U235 you get, and Kovarex ain't random. Finally, second-hand U238 from spent fuel reprocessing can be turned into even more U235, but that doesn't increase the 45:1 figure for reactors - because a batch of fuel rods takes 10 U238 to make; you're merely reclaiming that from spent rods. (Not sure how a prod mod would change _THAT_ tho...)
Anyway, I got here because the math looked off to me -- 1 _thousand_ ore per first-hand U235??? But that's right; the centrifuge takes a batch of 10 ore units per unit of U235. It's essentially more than just the centrifuge. It separates uranium from acid, tailings, and other impurities before separating the isotopes. So, one miner at 18 ore per minute makes 1.8 batches, and ~0.013 of those are U235. Which is enough to make 10 (or 11.6 with prod magic) times as many fuel rods, or ~0.15 of them. Each can run the reactor for 200 seconds, so you get ~30 seconds of reactor runtime out of a minute of mining operation; two would keep the reactor more or less supplied 24/7 while ore deposits last. Multiply that by 4 if you want to multiply the output of each reactor by 3, and you get the "8 mines" bottom line which Nilaus mentioned. (His treatment was a bit confusing because he didn't mention the ore shrinkage in the centrifuge, just that 0.7% (of the total output) is U235, and a bogus "18 per second" figure when he meant per minute. I hope my math is on point. I'm interpreting the "8 miners" figure as the requirement per plant of 4, rather than a figure per reactor block.)
I don't think Nilaus was suggesting that kovarex was useless, just that you don't need it straight away, and that you can implement it later
Funny, I grabbed the biggest uranium patch and put 177 miners on it and then had 52 centrifuges going. Once I watched and built this, I realize that is way overkill and there is no chance of my power plant ever running out of fuel, even if I duplicate it a lot. Thanks for the awesome info as usual!
Another great design that is very efficient in space and power. My nuclear is blueprint of choice until now has been the KOS nuclear plant. It is modular, but not nearly this compact. Thanks for the fresh look and design!
Excellent design and it seems you came to lot of the same conclusions as I did when doing my nuclear setup (although I never did the miners per reactor calculation, that is kinda crazy).
I also ended up with a 2x2 reactor setup (since you've got most of the neighbor bonus by then) and pulling water from one side without circuit controls or steam buffers. I sometimes add circuit control when I run these before unlocking kovarex but just adding a single storage tank is sufficient to monitor steam levels. Full steam storage is totally unnecessary.
It is good to see such a useful design getting highlighted. Cheers!
argument for steam storage:
unexpected power surges (or not looking at power consumtion) that enables you to put more steam turbines than the heat exchangers can usually handle thus allowing for a compact battery
This reactor design produces 480MW(th) out of which 464MW(e) is used to produce electric power. A4W reactors on US Navy Nimitz-Class Aircraft carriers produce up to 550MW(th) each (1100MW total), of which only about 100MW is used for electricity production (wikipedia), the rest of the capacity is used for ship's propulsion, plus some extra capacity for safety margin and sustaining the aircraft carrier with one of the two reactors down for maintenance or recovery from a casualty.
This design uses reactors that, unlike A4W reactors, operate at 100% capacity at all times. When operating A4W reactors, we typically maintain power as low as practical for the ship's operating conditions. Let's just say with both reactors running and normal day-to-day operations when in a mission area, they're typically operating at less than 50% each. The reason for using both reactors when one could carry the load demanded of both is to have operational flexibility if one reactor needs to go offline for maintenance or in response to a casualty.
Factorio reactors also use fuel enriched to... exactly 5% (1/20 parts U-235, 19/20 parts U-238) whereas the enrichment of navy nuclear reactors is... I forget exactly but it's legit well over 90%. It may be 98-99% even. This allows navy nuclear reactors to load enough power-production capability to last a ship for literally decades.
Someone with a lot more dedication than me could probably calculate the mass of U-235 consumed in 200 seconds to produce... well... actually, the fact that factorio reactors have this neighbor bonus thing kind of throws off the calculations. Why does a separate independent reactor adjacent to another reactor cause the fuel to produce more usable energy?
Maybe the neighbor bonus is modelling a larger combined reactor instead of many smaller adjacent ones. In a larger reactor core, the otherwise wasted neutrons could be used.
Thanks! I've just started Factorio, and I'm considering entering the nuclear age, I appreciate the beginner's tips as well as the more advanced guide.
Your content is sincerely appreciated. After finally building the nuclear plant, I can relax my worries regarding power consumption. Phew!
I just started generating power w/ reactors and turbines. The emperor would be proud of how many las turrets I have built.
I need to pull the blueprint, but I spent around 20 hours one time a year or so back trying to design the perfect nuclear power system. the end system used 6/8 reactors, double heat pipes, tanks for water and steam but not for stream storage because there was exactly 0 fluid dynamics checks because everything was moved by pumps. the water came in through the middle, between the turbines and the reactors, and across the bottom. if you flipped the blueprint 180*, you could stick another one on the bottom and get a slightly higher bonus, and if you put them side-by-side they could share heat to speed up the initial heating process, and would share a mutual water input stream. the pumping system could handle just less than 3x8 reactor setups horizontally before it would start to have problems getting enough water into the heat exchangers at the back. the last heat pipe held a constant 525 degrees when the system was fully hot and fully utilized, which is why I chose to stick to 6/8 reactor systems. can't get the heat out far enough to need any more than that. I think that's the earliest hard stopping block. when I say 6/8 reactor setups, I mean 8 to start untill all steam tanks are full, then take the bottom 2 away to leave 6 which maintain enough output to get 1.2gw steady, but built with enough turbines to do just about double that because that's what the system should be able to handle since it's duty cycle is only about 50%.
the best and only argument for steam storage is the day/night cycle and solar panels. Free(ish) power, and you definitely shouldn't need all of your reactors at full steam during the day if you aren't relying solely on them for your whole base's power. why not bank the rest of the steam for later? it's only a few tanks. pump in pump out and you won't have to worry about fluid dynamics checks. sure it's a bit bigger but, like, is your map size limited or something? just tack on a tank field connected by turbines to burn off the overflow in times of very high usage.
obviously I used bots to deliver the fuel. I mean, who would even consider trying to belt that stuff in and out. lol :P (I kid)
getting water in was always the problem, but can be overcome with proper placement and delivery techniques (pumps mostly)
All in all, that's a great guide though. well thought out and reasoned and the design is great for a 'bases first nuke field' setup. the math may work out to being able to do about 1 reactor to 8 miners, but I'd like to see that actually work in practice. what would the power consumption of 8 miners and (I think) 8 centrifuges with 8% productivity modules be compared to the power output of a reactor? what's the real power profit there? or is that just another really roundabout way of banking solar power when you think about it?
@Tommy Taffy has been a long time since I was a factorio master, but, when I designed this fluid dynamics were far more rudimentary than they are today. at the same time, it's likely that pumps bypass the standard fluid dynamics that seek to balance population and flow between connected pipes and fluid containers. pumps move fluid from one vessel directly to another with no mind to balancing or viscosity. if you chain pumps, or pump->vessel->pump, once filled, will flow at pump speed throughout the system (minus any fluid used or forked in previous steps). fullstop.
the gain in flow rate and fluid availability in deep systems is not negligible. try it for yourself.. chain pipes, underground, tanks... then do pump/tank/pump and see how much faster the end tank in the pumped system fills over any other layout.... then to prove it to yourself, in the middle of the pump setup, switch 3 pumps for 2 tanks and put 3 in a row in the middle of your loop, or connect with pipes instead, and see how just 1 fluid dynamics calculation substantially slows the rate the end tank fills.
I'm going to dig up that blueprint... I'm sure I still have a vanilla world from 3 years ago with it built somewhere.
I get it and I knew about the abundance of fision material since the first time I played the game (because I can read). I still don't like running my power plant at 1000°C.
Here is why I disagree with your perspective about steam storage:
One: steam storage is eeeeeasy. Just slap some tanks down in the free space along the sides of your boilers. The design difficulty is zero, and compared to the cost of the rest of the parts, a dozen tanks is basically free.
Two: peak capacity being higher than sustained capacity is useful for e.g. situations where you are transitioning from heavy solar or have lots of laser turrets or other situations where your power consumption needs might be "bursty".
Three: reactors which are hot (>900C) EXPLODE if they are destroyed by damage, with the same effect as a nuke going off. Nothing like a few biters getting into your power plant, and then the entire thing literally blows up and you have to rebuild all of those extremely steel heavy structures on a factory with no power.
Four: It adds resilience to brief outages. If I need to disconnect the water supply to re-route pipes, I can do that, and it will run for 10 minutes.
Technically, storing heat directly in pipes is more space efficient (500MJ/tile compared to steam tanks at ~269MJ/tile), but just building more heat pipes in random places is ineffective because as you point out, heat transfer becomes a bottleneck.
I disagree with your perspective about circuit-limited reactor designs too, because they are very simple to implement well. I homerolled a controller design which is 3x4 tiles, contains only 6 combinators, and it's both jam-resistant and idiot-proof. It's the difference between waiting until kovarex to spin up nuclear fuel for your trains, and starting it as soon as you can.
You make good points. I guess it depends on the situation: a big base with low UPS means not using tanks will make the game run faster, because it doesn’t deal with the fluid calculations involved, and a base on a limited-resources-run is way better off implementing efficiency measures.
It ultimately comes down to preference and convenience.
Well, if bitters chew your nuclear reactor, you have something very, very wrong with your base defence anyway.
I do agree that steam and circuit based nuclear power are simple to build, and they are scalable. You can easily build a 480 MW power plant (that's 4 reactors) even if you only need 60 MW right now (it's usually the power I need when I start looking into nuclear power). You won't use that much fuel, stockpiling for the future, and you won't need to deal power for a while.
@@giovannidueck9094 Megabase skip nuclear power entirely and only use solar. Solar have no UPS cost. Nuclear power can't beat that, with or without steam tanks.
Regarding point 3: With the addition of the spidertron, this has become a non-issue. Simply building the reactors on an island that is only accessible by spidertron makes them completely invulnerable.
Doesn't really matter once your Kovarex process is up and running, as you basically get unlimited nuclear fuel. Get a couple uranium patches being mined and processed and it's pretty much impossible to be short on it.
Thanks so much. I have recent been introduced to Factorio and I loving it. Your video are very insight. Much appreciated.
Great guide! It's really surprising how few miners and centrifuges are needed even without kovarex. I always start doing on-site uranium processing as soon as I unlock the research. I don't use modules, so I go by a 10 miners : 3 centrifuges ratio. I usually start with a 2 reactor build because its cheaper to build early on and satisfies my power needs for a long time, and just the 10 miners and 3 centrifuges is enough to support it indefinitely. The nice thing about having a tiny uranium processing build is that you can put efficiency modules in everything and never have to worry about biters. Sometimes I will use steam buffering for the 2-reactor setup because you only need a couple fluid tanks (maybe 10 or 20) and the wiring isn't too bad. For the 2x2 build though its definitely not worth it, especially after you get Kovarex.
Lemme put it this way, I LOVE this design. It’s aesthetically pleasing, works brilliantly and is easy to put down. Then a friend and I started playing with the Space Exploration mod, and this became INSANELY useful. That reactor blueprint is as vital and as much a useful a tool for us in that game as train stations are, we decided.
I would cut the turbine area in half and put it on the lower side. Just to make it more symetrical
There is a good way to measure the actual power produced with an "electric-energy-interface" (you have to create it from console). When you set it to consume power, it has the lowest priority, same as an accumulator. You'll be able to see on the charts how much power is going into it. This is nice because it not only reveals flow issues but also includes all the pumps or whatever you use to keep the reactors running.
An interesting point is that even if you're committed to steam storage for your reactors, you probably only need to build your first reactor with steam storage. You'll be building a second reactor because you're about to exceed the capacity of the first one, so every other reactor you build can be running at full load without "wasting" power anyhow.
This is what i do, I rush nuclear to build a 2x2 reactor with removable steam storage, I use the steam storage function while my base is under 480MW and remove the tank section when i hit full load and paste down another 2x2 reactor when I need more power. I know wasting uranium isn't really a concern but i like the puzzle of it and like not wasting things.
I have two extremely efficient designs that "just work", a 2x2 and a 2x5, both of which fit nicely within Nilhaus' city block blueprints. They feature auto start/stop for power saving and peak power boost thanks to steam storage tanks, backup power for the pumps nearest to the water input (assuming you disconnect their wires), all without a single combinator or requiring any sort of tuning.
The 2x2 has a max boost power of 500MW, and sustained 464MW of a possible 480MW, and the 2x5 has a max boost of 1.6GW, and 1.37GW sustained out of a possible 1.44GW, so they are highly efficient too. I tested them rigorously to ensure they do not suffer from any of the common pitfalls that Nilaus mentions in this video. You can of course remove the steam storage if you don't care about fuel efficiency or all those nifty features.
They are available on the Factorio Forums here: forums.factorio.com/viewtopic.php?f=208&t=91639&p=521863
The point about larger reactor setups (2x3 and larger) is only about uranium-saving is very helpful! Now I can use my 2x2 designs peacefully without worrying that I missed something by not going to 2x3
Just got into factorio two weeks back and your videos have been such a great help! I really liked this video cause you explained your methodology. Keep the great work up!
Thanks Nilaus, you completely convinced me to not use steam storage.
My little 600k uranium patch with a fully beaconed kovarex should last forever. :D
Hey Nilaus, this was a definite master class. Factorio graduate studies vs your 4 year first degree. Btw it is so cool when things like your uranium non-scarcity math appears on YT and I can remember being in the twitch stream when the calculation first happened ((: Bravo!
EDIT: I’m SO used to seeing radial symmetry nuke plans designs in your playthroughs. My eyes almost can’t believe this version. That’s the real controversial opinion hehe
I have about 70 hours into this game (rookie!) and did not touch nuclear yet. That might change soon! Thanks for explaining this so clearly!
If you wanted to get that missing bit of water in your ratios you could use a storage tank for water for each set of heat exchangers and use two water pumps per storage tank to keep the water storage tanks filled. Gives you the missing water and would still be tileable.
What an awesome guide. It saved my macaroni factory from constant blackouts.
One thing I like to do is add steam storage tanks and wire the system to feed fuel cells only when the steam storage runs low. This way, I use the fuel cells efficiently.
34:11
@@littlebuilder8132then just build 1x1 reactor (lone reactor) setups if you think uranium is so plentiful that you don't need to be efficient.
I love Nilaus, but his argument here makes no sense. He is saying to build 2x2 for efficiency, but then don't build steam/circuits because uranium is plentiful so you don't need to care about efficiency. Can't have it both ways.
you really don't need to be efficient with uranium once you have an enrichment process researched and running. you lose on a lot of free power if you build just singular reactors @@beepbop6697
@@beepbop6697 you need 12 1x1 reactors to get 480MW from a 2x2 , 12 reactors is far less efficient even with unlimited fuel.
It is only tileable if we disregard water supply. If base needs several GW of power it gets real messy real quick. I prefer more spread thin design, it also allows for +3 adjacency even though it is not a big deal.
Also, it would be excellent addition to show the design in action, to demonstrate that it is indeed producing at maximum capacity. Especially after what you have said about difficulty of designs with fluids.
Thank you for making these!
I haven't ran into any lag issues concerning nuclear power plants yet, and because of this I have developed my nuclear power very well. I designed a nuclear reactor that is actually tileable, and as long as ample water is provided takes full use of neighbor bonus. It consists of a base 2x7 reactor that then can be made into a 2xHoweverManyYouWant in a row. Only caveat is it must be built on top of a lake filled with landfills, since it would be near impossible to supply water with pipes. In my solo world, I was able to build it before I launched my first rocket, and each tile generates I believe 2.2 GW, which is amazing since you don't have to worry about power and if you do use more than that you just slap another tile one it. I have also developed a very efficient kovarex refining method that involves using circuit network signals to regulate exact proportions of u235 put into the centrifuges. It does require logistic bots, but if they are available to you you can easily convert thousands of 238 to 235 to be used for fuel or, my favorite use, spidertron with nukes. Of course, using this would leave you with very little 238 for uranium ammo, but if you use lots of laser turrets like me its not much of an issue.
If its requested, I can post my blueprints for either, but I'm sure there are already tons out there. Also, my reactor design requires lotsss of landfill (it has to be built ON, literally ON TOP OF a lake) and isn't going to serve as a beginner reacter lol.
Space Age comes out next week and I figured I should probably learn how to go about nuclear power and the purple and yellow science packs. Even now, 4 years after this video was made, it’s still very useful - it’s hard to figure out the ratios. (Hopefully Space Age’s production rate display helps with this and Fusion Power)
just be carefull do not step over your 1000ºC heat pipes, it can melt your boots
Space Age just made the argument of no kovarex or steam storage stronger, as there are now miners with Resource Drain stat , which is a new seperate bonus from Productivity, but it goes down to 16% drain on regular miners (so 84% of the time the ore patch does not go down), and starts at 50% for Big Mining Drill, up to 8% (so up to 92% of the time, it makes ore not consume the area)
Quality on Reactors/Heat Exchangers and Turbines also are quite interesting but I see it mostly a way to reduce space and also further reduces fuel usage, but needs calculations if mixing different quality levels.
Also you can now read the reactor temperature via the circuit network, so you can set it to insert (or empty the fuel cell) only above/below certain temperature, as heat pipes will continue to work as long as they are above 500c and temperature can go to double that. (so if it's a reactor setup , but only using partially, the temperature gets used up slower, so fuel gets saved by cooling down to let's say 600, and since they also added a way to read fuel inside reactor now, it never inserts more then 1 if you set it so.
Also adding more pumps exchangers and turbines will be interesting when quality comes into play (changing ratios if quality is mixed) the but the new fluid system makes pipe length irrelevant (up to 320 length) as the length of pipe is one giant storage tank now.
Kovarex is still helpful for other things, as nuclear missiles cost 100 U235 now
i have a 2x80 power plant for almost 20Gw. Tricky part is finding a perfect lake to build on it
Thank for the fluid info, I will downsize a lot of storage.
I managed to build a 2 GW reactor (14 reactor cores) and got it working flawlessly. No steam problems, no heat problems, no water problems, ran at full capacity. I figured the efficiency would be nice, especially in the late game, but any more than 14 reactor cores and I start having issues that I can't resolve.
The problem with a 2 GW reactor, as mentioned in the video, is that it's absolutely huge. I had like 5 of the damn things, best place I could find only fit 2 of them, the other 3 were just placed wherever I could find space. I think smaller builds probably would be better, despite being less efficient. I'm going to try building a compact 1 GW design for my next run.
8 reactors in a 3x3 grid, with one of the sides removed is actually the most efficient in terms of the neighbor bonus. Use long handed inserters to put the fuel in, and you're good to go.
I actually sitted trough this whole video. This was one of the best masterclass videos yet. More of these!
Grammar police: You *_sat_* through. But yeah I'm really glad he did this one, one of the most useful topics so far.
I'm not sure if that's what the nuclear fuel recipe has always been but I swear I remember one of the main reasons you needed Kovarex was because of the sheer amount of U-238 that you'd have to store which was why so many people wanted an incinerator of some sort for the game.
Ah, I always wondered why he invented that.. It would make sense exactly in the situation that you describe. And it would also explain why we always have way too much uranium patches by default.
the dlc expansion has a recycler, from what i understand it cant un-refine raw intermediates so it should return 25% of the u-238 thrown in
I don't have the funds to support, and thereby show my approval at this time. But I am truly grateful that I can benefit from your knowledge, effort, and passion like this. I hope that my gratitude counts as a small token. (But I also hope that you are getting worthy financial support as well). It's like the feeling when you see a street performer who is breathtaking, but you have no cash on hand. You feel almost like you've taken from them as you walk away, without having given back.
Thank you. Don't worry about it :) I provide this content for free and hope that a small percentage want to and is able to support. I do not want any paywalled content, so just enjoy it :)
I saw a Reddit thread where someone showed off their 2x5 build. They showed screenshots that demonstrated it was indeed running at peak efficiency, and it was absolutely beautiful, but my gods, I cannot imagine how many hours they spent on it. Unless you intend to just copy that person's blueprint, I would highly suggest taking Nilaus's advice here.
or use water well pumps (offshore pumps buried under landfill), provided you have access to a big enough lake, it solves the trilemna of heat/steam/water connection : just feed the heat exchangers lines directly with water well pumps and the rest will get far easier to deal with.
DL the badass BPs if you don't feel like doing it yourself. There's a 2.6 G All circuited up and it works perfect. So yeah, figure out your own designs but once you go big don't bother, it's already been figured for you, unless you just like that kind of challenge. Very instructive video for me. Thanks for making the video.
Steam storage and overbuilding turbines is good for dealing with surge demands (like a seriously oversized laser defense grid)
Larger nuclear reactor setups save space overall.
Start mining Uranium right away, save enough to start enrichment(not much honestly), then build your reactor fuel supply. When enrichment comes you can start right away and hit those milestones for nuclear weapons and nuclear fuel for your trains.
My entire factory consumed less power than a single reactor outputs (b/c solar) but I wanted better efficiency so built a 2x2 reactor anyway.
No way in hell am I wasting all that fuel running full time. Definitely worth the circuit effort... though I needed to make a T-flop train circuit to keep accumulators charging on only solar. That part was a mess lol
"if you are not an absolute master in designing these, don't do it, its not worth it"
Its a designing game, you design something and work out to increase its efficiency.
Yea, and how are you supposed to become a master without trying?
Well, I think it depends on what your goals are. Some people like to make the larger setups just to do it and because they enjoy the design process; others are more practically oriented, so beyond a certain point if it doesn't actually further benefit the functionality of the factory it isn't worth doing.
@@strategicsage7694 If you do not understand how a setup/blueprint design works; you shouldn't use it. Thats almost all the fun of this game. I don't know a single person who enjoys factorio just by copy pasting some blueprints off the internet and launching a rocket.
@@kitsukeita Where do you draw these people you know who play factorio from? That's certain not what I observe. I see a lot of channels where the biggest request is for the blueprint downloads. A number of people specifically say it's because they can't be bothered to make their own designs and/or just want to get past whatever aspect is causing them trouble. People play Factorio for a lot of different reasons, but there's definitely a sizable group of people who are more interested in the solution (blueprint) than the understanding of it or the design process.
@@strategicsage7694
People might ask for blueprints for various reasons, maybe they want to assimilate it in their own design or they want to further improve it or they just like it more than their own design; but if you do not understand what the blueprint is supposed to be doing & how, i see no point in playing; might as well copy those small starter base prints from the internet(meant for advancing faster in modded games), hand craft the components/buildings, launch the rocket and call it a day.
My data of people comes from my steam friends/factorio steam groups/Bobs,Py + some other mod discord channels.
Who was hoping the thing would blow up when Nilaus said he had never seen it before :)
and you should add accumulators as well to store any excess energy since it was built to run constantly in this form. That would make it more viable and unlike steam storage this would allow it to run without losing any energy overflow.
@Tommy Taffy Ah ok still 20 loss is better than 100 :)
I use steam storage with additional turbines as a means to store energy for turret spikes. I also setup an alert system so that once it is below a certain point(power use exceeds actual power production) I can go add another module to the power plant.
you can also use the steam storage to prevent over heating of your reactors : nuclear plants heat up to 1000°C no matter the consumption, any excess heat is wasted. If you setup things so that you only empty your reactors when steam storage goes bellow some level, and you only fill back when emptying, you will minimize fuel waste.
I happen to be a very inefficient player, so my factory backs up a lot and fluctuates heavily in consumption.
Steam storage always felt right to me as they are essentially “better accumulators” but the performance argument makes sense. Exchangers produce slightly more fluid than the turbines can use which ensures the pipes are always filled.
I tend to just go for “max efficiency/perfect rate matching “ when the better idea is to just overbuild everything as nothing is really wasted in this game
I wish I had seen this video a month ago. I have a space exploration megabase krastorio 2 game going right now. On my main planet I ran out of all nearby uranium sources and had to start importing it from other planets because I actually did my Kovarex too much/fast/whatever and completely drained all of my bad uranium. I had this huge panic when I saw my nuclear fuel generation had hit zero and it wouldn't restart because I didn't have all the needed materials. Thankfully, I had stockpiled centuries worth of fuel to keep the base defense lasers going long enough to start getting cargo rockets of uranium sent back home.
TL;DR - Thanks for the math at the end, super cool stuff I wish I had known forever ago.
For people watching this after 2.0, making these fuel efficient became trivial now that circuits from reactors emit a fuel signal and a temperate signal. A very basic circuit can make 1 uranium fuel last many orders of magnitude longer.
link to template tutorial?
@@Klorel123 It's not letting me post a link for some reason. So here's a summary of what I did. You can DM me if you want a link to the blueprint.
It's very straightforward. You just make each reactor output temperature and fuel signals. Then, for each reactor, you just have one decider check the temperature signal and output 1 on H if the temperature falls below some threshold (I chose 650°C); and you have another decider check the fuel signal and output 1 on Uranium Fuel Cell if the fuel from the reactor falls below 1. Then you connect both of those deciders to the input of an arithmetic controller and have that output H AND Uranium Fuel Cell as R.
Attach the output of the arithmetic controller to the inserter that puts fuel in the reactor and check Enable/disable and set the condition to R > 0 and override the stack size to 1. Do this for all four reactors separately.
This makes it so that the inserter will only load 1 fuel cell into its reactor when the temperature is below 650°C AND there's no burning fuel in the reactor.
@@aloric711 I'm just trying to learn logic in the game, can this not be done with one decider using temp AND fuel cell conditions to output 1, then the output of decider to the inserter? so need need for the arithmetic?
@MrRiverz88 A decider can only check one condition.
Using those numbers, we should only need around 2 minuers per reactor. Either the ratios you presented aren't correct, or else I made a mistake in my calculations somewhere. Can anyone help with this?
So we have:
In the miners:
Uranium ore per miner per minute = 18
Uranium ore per miner per second = 18 / 60 = 0.3
In the enrichment process:
Amt. of Uranium Ore per unit of Uranium 235 = 1323
In the assembling machine:
Amt. of Uranium 235 per unit of Uranium fuel cell = 10/116 ~= 0.08
In the reactor:
Amt. of fuel cells per reactor per second = 1 / 200 = 0.005
Putting them together:
# of miners to constantly power 1 reactor = (1/Uranium ore per miner per second) * Uranium ore per Uranium 235 * Uranium 235 per Fuel Cell * Fuel cells per reactor per second
= (1/0.3) * 1323 * 0.08 * 0.005
= 1.764
This means we need less than 2 miners per reactor, as opposed to the ~5.6 miners mentioned in the video.
I think he said it wrong, it's 7.6 miners for the block of 4 reactors. Nilaus often misspeaks but then he is not a native English speaker. (Although it really irks me that he calls "Sulphuric acid" "Sulpher" ALL THE TIME! ) Doesn't detract from the amazing builds and tutorials he produces
Hmm, this is interesting. I didn't know *just* how abundant uranium was, though I was aware it was ridiculously abundant. I enjoy designing and have made some ridiculously large power plants. I've done a 2x16 model and I think it's the best way to learn about water and flow mechanics. The amount of pumps just to increase flow for the water I had to use was staggering and the steam required a lot of attention too. I've ended up with hundreds of thousands of Uranium, both shiny and dark, and I guess I never really thought about it beyond knowing that between steam storage as batteries and efficiency from the amount of reactors I was using that I was cutting down on my uranium consumption significantly.
I'll probably keep doing that. Definitely once I get to the point where I need more than 4 reactors, the design problems involved with nuclear is very entertaining :D.
The reason I use the refining process is to consume the 235. My biggest problem with nuclear power is the lack of ability to delete items. In my first time using nuclear power I set up 20 steel crates to hold all the 235 and once they filled up my centrifuges stopped pulling out 238.
The twitch chat was .. wow, yep. I wont bother checking anything else you yours build! Great job, team. -Sean.
Reminder to disable snap to grid on the reactor builds. They don't play nice if your blocks are aligned to the rails, but not to the very specific grid used to make it.
yiu might not need koverax but if you don't have have it you have to store U238 somwhere. Also if you don't use koverax to feed your reactors it will take even longer to get the U235 you need to start koverax since now your also using up some of it to run your reactora
This was awesome. I think I'll finally get into nuclear now haha
Your best yet, Nilaus! I especially appreciate the theoretical background. I see your point regarding the calculations, but frankly I also see it as a challenge to do it better... :)
good design but inefficient. You only put in enough 235 to heat the exchanges. then take the 235 back out. Storing the Heated Steam is the most Efficient Method Period. All you have to do is create a simple circuit to register the 500/750/900c or what ever your mod is. Once the Exchangers hit target temp filter back out the 235, when target reaches below like 30% of max temp filter back in.
Thanks Nilaus, if I ever come back to factorio I will know for the future on how to do a nuclear set up. My first attempt at it worked but it wasn't perfect and for some reason I NEEDED steam storage for the damn thing to be stable production. Next time I'll use your design. Love your videos and designs bro!
Just use 1 more pump and connect it to water pipe network, no need for another separate input, pipes can transport more then 1pump worth of fluid.
I know this is an old video, but I still kinda want to discuss some things if possible.
First of all while I do agree with many of your arguments regarding steam storage and big reactors, but they are based on the question of whether you need to save fuel or not.
So I feel a bit dissatisfied about the argument that you never need to build bigger than this one based solely on the fact that fuel is cheap. 2x1 would make for an even simpler build and probably even easier to tile - it's not more complex than a typical coal burner setup, so why not use it? I think saying that 2x2 is all you need needs a little elaboration, on why not smaller.
Personally, I think the question regarding the size of the reactor is about what you feel comfortable building yourself. On my last save I personally went for 2x3 reactor supplied by bots with nice symmetrical design, meant to be stamped all over the lakes. 2x3 is not much more complicated than 2x3 - requiring 80 heat exchangers and 138 turbines, and it outputs a nice round number of 800 MW
Next I do kinda disagree with you design process of limiting your reactor output by water pumps, like other comments have mentioned. Water pumps are really not that expensive, and sometimes building more of them just produces a simpler design with less piping involved, especially if we are doing landfill design that's meant to go over lakes - you can just set up a pump for each row of heat exchangers, and leave it at that.
So I personally overbuild my water pumps and turbines a bit for the sake of simplicity and symmetry - like for example my 2x3 design actually has 144 turbines instead of 138, because it consists of 8 separate sections with 18 turbines each, and I want each being able to give maximum output. I can connect them together and remove some of the turbines, but honestly I don't see the need for that and the current version is cleaner and aesthetically pleasing. The setup has also been tested, and it does produce exactly 800 MW at maximum capacity like it should.
There is of course a consideration that you usually want to have more power available than your base consumes, so having your design limited by water pumps is probably pretty much irrelevant in the long run, I guess our difference here is that I prefer to round up and you opted to round down in this case.
Agreed. If the argument is that uranium is so cheap/plentiful then why not just build 1x1 (lone reactor) setups? Anything larger is for the purpose of getting the neighbor bonus which *saves fuel* .
So on one hand he says you need 2x2 for the efficiency, and then on the other hand limiting the water (and no steam/circuit) because uranium is cheap and plentiful so just ignore the inefficiencies.
Can't have it both ways.
I can think of a reason to use steam storage on your first pre-kovarex reactor and that is to save U235 to get kovarex started ASAP. The less you use on fuel cells, the sooner you can have effectively unlimited U235 to make trains go fast and nuke every forest in your way, which is the proper use of nuclear power.
Thank you so much man for this tutorial, I was having so much problems with my last nuclear setup and I was running from here to there trying solutions to patch them but when i think i had fixed one then another came up. This is a like and a new subcription
You rock N! This is what i needed to reacquaint myself with nuclear. stopped playing for awhile after 0.12 and wanted to refresh after 1.0 and all of this free time. you are excellent at explaining.. gotta say the statement '1 fuel to 200 reactors' is a nonsense statement but it sure is attention grabbing :) I've noticed in the past that there isn't any point in steam storage - except the challenge *grin*
For others, here's another way to break down the math behind "uranium-235 is more abundant than you think" is in terms of time.
There's just under a 1% chance of getting a uranium-238 each time uranium is refined, which can also be written as a 1/100 chance (to be more specific, a 7/1000 chance). Uranium refining takes 12 seconds. If you had 10 centrifuges all getting a steady supply of uranium ore, they would refine 100 uranium in two minutes; if you scale back to a more conservative 5 centrifuges, they'd refine 100 uranium in four minutes. That means on average, with five centrifuges, you'll be getting a piece of piece of uranium-238 every four minutes, five minutes to lowball it if you're feeling unlucky.
A single piece of uranium-238 produces 10 nuclear fuel cells, which in total will run a reactor for 2000 seconds, which is 33 minutes. So, once that first piece of uranium-238 pops out, it's going to provide you more than 30 minutes of power, and you want to get another one to make more fuel cells before time is up. But within that 30 minutes, you'll have around 6-7 "almost a sure thing" chances. And this is all without modules to increase speed or productivity, and on just five centrifuges.
The odds are ridiculously in your favor that you'll get at least one uranium-238 in time, likely more than that, and of course as you keep playing the stockpile safety net will keep growing. The catch then is just making sure you've set up your output lines in such a way to get the 238 to the assembly machine making the fuel cells in a prompt manner, but that's no biggie.
Im just looking at setting up my first reactor but didn't he say 200 seconds?
@@JonnySparta one fuel lasts 200 seconds, but a uranium-238 produces 10 fuel cells.
@@DrakeyC Ah I see, my bad :) i did kind of figure anyone writing a comment like yours probably did know what they were talking about
@@JonnySparta No problem :) I'm not an expert at the game by any means, I just enjoy math.
I have been converted. Concise arguments, wonderful class.
If I wanted to conserve fuel, I'd build solar panels. I design ridiculous nuclear plants because I find them interesting to design.
My first reactor is always a single reactor using steam storage. Then my second is a 2x1 or 2x2 depending how my supplies are. Then I deconstruct all of my reactors and build a 2x3. I'll try just sticking with 2x2 from now on. Thank you.
Really great with the calculations of miners needed, and as a result of that, there is no need for more than 2*2 designs. Hope there will come a blueprint for Kovarex at a point too.
"steam storage is useless"
Bob's and angels: are you sure about that?
To be fair, ive been watching for ages and actually took notice to check it i was subscribed and i wasnt. I swear i subscribed a while ago, sorry Sir! Love the content it really is absolute mastery
Great design, great video, great job. Thanks all those involved.
Something one may find interesting is in real life the contetration of u235 is about.7% and the amount of u235 in reactor grade fuel is 5% the amount of u235 in a fuel rod in Factorio is 1 U235 and 19 u238 which is 1in20 exactly 5% so the fuel is actually reactor grade also sulfuric acid is actually used in uranium refining
That Maths just blew my mind.... How can it be so low?!?! :D
How much excess of the "Waste Uranium" does that "8" Miner process produce per minute after producing Fuel? If I am not converting it to Ammo, I would possibly need huge storage values until Kovarex Refinement comes online. I might try the maths on that myself.
Ok I'm new to Factorio but been playing Satisfactory for a while now and there is some mechanics or way it work that you probably assume everyone knows that new player don't. Things you don't explain here is :
1. How many or how much sulfure do you need to mine the ore.
2. The build leading up to the nuclear fuel. How much constructor do I need for that build to work ?
3. What do you do with the waste product ? Uranium 238 and the fuel waste ?
4. What does Kovarex do ?
1) One sulfuric acid (wish he would stop calling it sulfur, which is a different ingredient) per mined ore
2) Not sure what you mean here. It takes blue science to research nukes, at that point you will have all the necessary assemblers. (mark 2 assemblers and oil processing, specifically)
3) U-238 can be used for very powerful ammo and for kovarex (see 4). The fuel waste can be reprocessed back to make more U-238.
4) Kovarex is a way to convert U-238 into U-235. It takes 40 U-235 and 5 U-238 and turns it into 41 U-235 and 2 U-238.
“You don’t need a lot of uranium”
Me after searching for uranium for an hour: 👁👄👁
I went with a circuit and steam storage design because otherwise of dozens nuclear power plants would have been sitting at 1000 c degree all the time (maximum capacity) and still consume fuels.
All other energy demands are fulfilled by solar arrays during the day.
Would you rather have fuel cells be consumed all the time? Or occasionally be turned off?
It is kind of like having regular boiler consume coal/tree/etc... all the time. Lot of wastage from not utilizing all of the boiler output (1.8 MW) and miniscule power to move items.
Going large scale late with multiple reactors I can see how not using steam storage actually might end up beneficial. But for your first reactor using steam storage and a simply inserter circuit costs you basically nothing but saves a lot on actively consuming infrastructure you need to support your reactor, especially since when you migrate to nuclear power or the next reactor level you will most likely not yet fully utilize the power available to you and/or have some solar power in place.
Fully agree on Kovarex however. That's something you can put off until you want to fuel trains with nuclear fuel.
Thanks again for some more great content. Really appreciate your hard work and clarity on these guides. So helpful.
Can I ask a question, though? I don't understand how you worked out that centrifuges require an average of 1.3k ore to be mined to produce only 1 piece of U235. My calculations work it out at 132.3, so that's makes processing of ore into U235 10 times more productive than you say it is (30:05).
This being the case, it actually make your point that you don't need to bother with steam capture or kovarex 10x more potent!
What am I not understanding? Am I missing something? Thanks!
---
Here's my math, just in case.
U235 mining yield @ 0.7% = 0.007; 8% production bonus on centrifuge = 1.08 centrifuge yield; the combined yield = mining yield x centrifuge yield;
=> combined yield = 0.007 x 1.08 = 0.00756
∴ Combined avg. yield = 0.00756
Calculate the inverse to switch unit to U235 rather than ore; => 1 / 0.00756 ≅ 132.3 : 1
∴ Avg. ore needed to produce one U235 is approximately 132
the uranium centrifuge recipe takes *ten* ore to produce *one* bit of uranium. it takes 142.85 (1/0.007) cycles of the recipe to generate one unit of u-235 on average, so 142.85*10 ore, less the gains from productivity, which brings it down to around 1.3k ore.