Every Factorio nuclear guide repeats the same fact: one reactor makes 40 MW. Almost none of them show the arithmetic behind the ratio that follows from it, and none connect that ratio back to the uranium miners and centrifuges that have to feed it. That gap is where new nuclear bases actually fail — not at the reactor, but three steps upstream, where a starved centrifuge quietly stalls the whole chain.
This guide walks the full path from ore to megawatt with the exact numbers at each step, sourced from the game’s own recipe data, plus the one timing decision (when to flip on Kovarex enrichment) that determines whether you spend the midgame drowning in U-238 or running out of it. Verified against Factorio 2.0 / Space Age data — the reactor, heat exchanger, and turbine numbers themselves haven’t changed since version 0.15, so this ratio is safe to blueprint and forget.
Quick Start: Build This Order
If you want the setup running today, follow this sequence — everything after this section explains why each step is sized the way it is.
- Research Nuclear power, Kovarex enrichment process, and Nuclear fuel reprocessing before you commit any belts.
- Place 4-8 electric mining drills on a uranium patch, piped with sulfuric acid.
- Feed them into 1-2 centrifuges running the base Uranium processing recipe.
- Once you’ve banked 40 U-235, dedicate one centrifuge to the Kovarex loop and never let it run empty.
- Build 1 reactor, 4 heat exchangers, and 7 steam turbines, fed by a single offshore pump.
- Automate fuel cell assembly (1 U-235 + 19 U-238 + 10 iron plate → 10 cells) so the reactor never idles.
- Scale to a 2×2 reactor block only once your factory’s power draw actually exceeds 40 MW — see the scaling section below.
The Only Ratio You Actually Need: 1 : 4 : 7
A single nuclear reactor produces exactly 40 MW of heat, and every part of the chain downstream is sized off that one number [1].
A heat exchanger absorbs 10 MW, so a lone reactor needs exactly 4 heat exchangers to consume its full output — no more, no fewer [1]. Each exchanger converts that heat into roughly 103 units of 500°C steam per second. A steam turbine can only swallow 60 steam/second, which means one exchanger needs 1.718 turbines to fully drain it [1]. Multiply that by 4 exchangers: 4 × 1.718 = 6.87, which rounds up to 7 steam turbines. Round down to 6 and you’ll bottleneck the exchangers with backed-up steam; round up to 8 and the extra turbine just idles.
Water is a non-issue at this scale: 4 exchangers pull about 41 units of water per second combined, and a single offshore pump supplies 1,200 units/second — enough for roughly 11-12 heat exchangers before you need a second pump [1]. One reactor also burns exactly one fuel cell every 200 seconds, regardless of how many neighbors it has, which matters for the mining math below [1][4].
Feeding the Chain: Miners Per Centrifuge
This is the step every other guide skips, and in practice it’s the one that actually causes new nuclear setups to stall — I’ve seen more “finished” reactor builds bottleneck on a starved centrifuge than on anything downstream of it. A base electric mining drill on uranium ore produces 1 ore every 4 seconds — 0.25 ore/second, unmodded [7][8]. A centrifuge running the standard Uranium processing recipe consumes 10 ore every 12 seconds, or 0.833 ore/second [3].
Divide those and you get 0.833 ÷ 0.25 = 3.33 — so 4 uranium miners fully saturate 1 centrifuge at base speed, with headroom to spare. Undersize this (2-3 miners) and your centrifuge idles between ore deliveries, which is a quiet, easy-to-miss reason a base that looks finished still runs out of U-235.
Sulfuric acid is the other half of the equation: uranium drills consume acid 1-for-1 with ore mined, so 4 miners need about 1 acid/second [7]. One unmoduled chemical plant produces 50 sulfuric acid/second [9], which is enough to fully supply roughly 200 uranium miners at base mining speed — acid is essentially never the bottleneck unless you’ve built a genuinely enormous uranium field. If you’re running mining productivity research or speed modules, this 4:1 ratio shifts; treat it as your unmoduled baseline and adjust up as your mining speed increases.
When to Flip the Kovarex Switch
Raw ore processing is brutally unfair by design: each 12-second cycle has a 99.3% chance of giving you U-238 and only a 0.7% chance of U-235 [3]. That’s not a balance oversight — it’s exactly why the Kovarex enrichment process exists. Kovarex converts surplus U-238 into more U-235 at a 3:1 ratio (the recipe consumes 40 U-235 and 5 U-238 to produce 41 U-235 and 2 U-238 over 60 seconds; the 40 and the 2 cycle straight back in as catalysts, so the net result is 3 U-238 in, 1 U-235 out) [2].
The practical question is when to start it. You need 40 U-235 banked before a Kovarex centrifuge can run its first cycle. At the base 0.7% drop rate, the expected wait is roughly 5,700 processing cycles — around 19 hours from a single centrifuge running alone [3]. Nobody actually waits that long: run 4-6 centrifuges on raw ore from the moment you unlock uranium processing, let them run in the background while you build the rest of your base, and 40 U-235 shows up well before you’re ready to build reactors. As a general guideline, don’t dedicate a centrifuge to Kovarex until you have that full 40 banked — pulling the trigger early just stalls the loop on an empty input.
Here’s the mechanism most guides miss: Kovarex isn’t really about feeding reactors. A single reactor’s fuel demand is tiny — one fuel cell (1 U-235 + 19 U-238 + 10 iron plate, crafted 10 at a time) every 200 seconds works out to about 0.0005 U-235/second and 0.0095 U-238/second [4]. A single raw-processing centrifuge already produces U-238 at roughly 0.83/second — wildly more than any small reactor cluster needs. The real problem Kovarex solves is the U-238 backlog: without it, U-238 piles up in centrifuges and storage until the whole processing line stalls, which throttles your already-slow trickle of U-235 along with it. Kovarex is a release valve for that surplus, not a fuel-cell requirement.
Depleted fuel cells aren’t wasted, either, once you’ve researched Nuclear fuel reprocessing: 5 used cells convert into 3 U-238 over a 60-second centrifuge cycle [5]. It only recovers U-238, never U-235, so treat it as a minor efficiency bonus, not a real Kovarex shortcut.
Scaling Up: the Neighbor Bonus Math
Reactors get more efficient the more of them you build, because of one rule: every reactor gains +100% heat output for each adjacent reactor, counting only orthogonal (N/S/E/W) neighbors — diagonals don’t count, and the bonus boosts heat only, never fuel consumption [1].
A 2×2 block is the standard next step because every reactor in the square touches exactly 2 neighbors, giving each one +200% heat: 40 MW × 3 = 120 MW per reactor, 480 MW for the block. Scale the exchangers and turbines off that total the same way as before: 480 ÷ 10 = 48 heat exchangers, and 48 × 1.718 = 82.5, rounding up to 83 turbines [1][10]. Fuel consumption doesn’t scale with the bonus — 4 reactors still burn exactly 4 cells per 200 seconds combined, or 1.2 cells/minute, since the neighbor bonus is a heat multiplier, not a fuel multiplier [1][4].
Past a 2×2 block, the pattern for longer 2-row strips is straightforward: interior reactors pick up a 3rd neighbor, and each one it gains simply adds another +100% heat, another 4 exchangers, and another ~6.87 turbines to that reactor’s share. Don’t hand-place a strip past 2×2 without a blueprint — I’ve miscounted a turbine row by hand more than once at this scale, and the exchanger/turbine count grows fast enough that a manual count is exactly where most players introduce an off-by-one error.
Which Setup Actually Fits Your Base
| If you are… | Build this | Why |
|---|---|---|
| New to nuclear | 1 reactor, 4 exchangers, 7 turbines | Smallest setup that can’t be miscounted; learn the ratio before you multiply it. |
| Casual, want it done once | 2×2 block (4 reactors, 48 exchangers, 83 turbines) | One blueprint, 480 MW, covers most mid-game factories without a second build. |
| Optimizer / min-maxer | 2×2 block sized to exact current MW draw, Kovarex running on 2+ centrifuges | Avoids overbuilding exchangers/turbines you don’t need yet; U-235 supply keeps pace with expansion. |
| Completionist | Full reactor row with reprocessing loop active | Squeezes extra U-238 out of every spent cell; only worth the centrifuge time once uranium mining is otherwise idle. |
Nuclear Isn’t the Endgame Power Source
Fission is the mid-to-late-game workhorse, not the final answer. Once you reach Aquilo, the fusion reactor and fusion generator combo outputs more power per building than a 40 MW fission reactor, scales fuel use directly to plasma output instead of burning a fixed-rate cell, and leaves no spent fuel behind [6]. Most players still import nuclear fuel to bootstrap early Aquilo operations before switching over — so the ratios in this guide don’t become obsolete, they just stop being your primary power source. For what comes after fission, see our Space Age endgame guide.
If you haven’t hit Vulcanus yet, note that foundries there lean on direct heat piping rather than steam power for some processes — worth knowing before you assume every planet wants a copy of this exact reactor block. Our Vulcanus guide covers the foundry side in detail.
FAQ
Should I build Kovarex immediately after researching it?
No — wait until you’ve banked the full 40 U-235 it needs as a starting catalyst. Dedicating a centrifuge to Kovarex before then just leaves it sitting empty, doing nothing, while your other centrifuges could still be stockpiling toward that 40.
Is 1 offshore pump enough for a 2×2 reactor block?
Yes, with room to spare. A 2×2 block’s 48 exchangers pull roughly 494 water/second combined, well under a single pump’s 1,200/second capacity — you’d need a much larger reactor row before a second pump becomes necessary.
Why do my turbines keep backing up with unused steam?
Almost always an exchanger/turbine mismatch from rounding the wrong way — building 6 turbines per 4 exchangers instead of 7 leaves steam with nowhere to go. Recount against the 1.718-per-exchanger ratio rather than eyeballing it.
Do I need reprocessing researched before I build reactors?
No. Reprocessing only recovers U-238 from spent cells, and reactor fuel demand is so low that U-238 is rarely the bottleneck early on. Research it later, once your uranium mining outpaces your Kovarex loop and you want to squeeze extra value from every burned cell.
Sources
- Tutorial:Nuclear power — Official Factorio Wiki
- Kovarex enrichment process — Official Factorio Wiki
- Uranium processing — Official Factorio Wiki
- Uranium fuel cell — Official Factorio Wiki
- Nuclear fuel reprocessing — Official Factorio Wiki
- Fusion reactor — Official Factorio Wiki
- Uranium ore — Official Factorio Wiki
- Electric mining drill — Official Factorio Wiki
- Sulfuric acid — Official Factorio Wiki
- Factorio Nuclear Reactor Calculator — factoriocalculator.blog
I've been playing video games for over 20 years, spanning everything from early PC titles to modern open-world games. I started Switchblade Gaming to publish the kind of accurate, well-researched guides I always wanted to find — built on primary sources, tested in-game, and kept up to date after patches. I currently focus on Minecraft and Pokémon GO.