Fixing Early-Game Power Shortages in Factorio | Steam Ratios and Recovery Steps

Blackout loops in the early game of Factorio—especially when you're automating red and green science—are the most common jam-up point new players hit. This guide cuts through power shortages by sorting them into three quick diagnoses: insufficient generation capacity, fuel starvation, and night-time demand shortfalls. You'll work through the fastest recovery sequence step by step.

The guide covers the essential maths: proper steam ratios, the 25:21 shift towards solar panels, and the key reference that 1MW requires roughly 23.8 solar panels and about 20 accumulators. It distils down to just the numbers you actually need. Early on, I hit this wall too—the whole factory ground to a crawl, research and mining both tanking—but once I zeroed in on coal supply and switched to prioritising the power plant, adding just one more steam unit got things running again in under five minutes. The trick is that power loss responds better to ratio-based thinking than brute-force expansion. This guide's main aim is to stop the early blackout where it stands and set up a power structure that won't choke later.

Target version, assumptions, and terminology

This guide's progression level and research requirements

This covers vanilla Factorio v2.0 at the point where you've got steam power running and are just starting to automate red and green science. Stone furnaces, boilers, steam engines, labs, and assemblers are beginning to cluster together—this is when total power draw jumps sharply. You're not yet on full solar, and it's the moment you need practice reading the power screen to understand what went wrong.

There aren't many techs to worry about at this stage. If steam power is already working, you don't need high-tier research to fix a blackout. What matters now is understanding the relationship between current generation and consumption. A steam engine outputs 900kW max per unit, making it an excellent reference point for early-stage output. An assembler consumes 90kW when working, so ten of them roughly match one steam engine's output. When you frame it in numbers, power shortages become trackable by ratio rather than pure intuition.

I also fell into this trap as a beginner: red and green science ticking along, but "mining and research are too slow to matter—why does adding more power plants help nothing?" The culprit was coal supply priority, not raw generation. Early on, misreading your existing setup trips you up more than missing research.

This guide also assumes Space Age DLC is not in use. Since Factorio 2.0 and Space Age both launched on 21 October 2024, the rules for power design can differ sharply between vanilla 2.0 and expanded/modded environments. Everything here is tuned for vanilla Nauvis early-game and will work as-is without add-ons.

Core terminology: fulfilment ratio, production, and consumption

Three terms matter most: fulfilment ratio, production, and consumption. In this context, fulfilment ratio means how much of the demand you can actually supply, expressed as a percentage. If you're at 100%, everything runs at normal speed. Below that, machines slow down proportionally—so the factory doesn't cut out suddenly; it feels like everything grinds to a halt gradually.

Production is the power your generation is actively outputting right now. A steam engine can max out at 900kW per unit, but it doesn't always hit that peak—it supplies just enough to meet demand. Consumption is what your machinery needs. When consumption exceeds production, the fulfilment ratio drops, and miners, assemblers, and labs all slow together.

A common mistake is treating accumulators as main power. They're not. Accumulators charge when there's spare power and discharge only when other generation falls short. In other words, they're your last resort. Vanilla accumulators hold 5MJ, with a max discharge of 300kW per unit. Perfect for bridging dips or covering the night, but not a substitute for underlying steam shortages. To replace one 900kW steam engine, you'd need three accumulators just for the discharge rate. That's why stacking accumulators early without addressing the root cause wastes time.

Solar works similarly: a single solar panel outputs 60kW max. It works during the day but vanishes at night. To run around the clock, you need accumulators. The vanilla baseline is 25 solar panels to 21 accumulators (from the Solar power|Power production/ja wiki). To sustain 1MW over a full day cycle, aim for around 23.8 solar panels and about 20 accumulators. Building that from scratch early is heavy, so ramping solar gradually while keeping steam as the backbone is more practical.

Power production/ja

wiki.factorio.com

Vanilla vs. Space Age/large mods: the differences

The key lesson here is: don't mix vanilla knowledge with mod knowledge. In vanilla v2.0, offshore pumps need no power to run. Plant one by water, pipe it out, and it keeps pulling even when the grid dies. So in early vanilla blackouts, the frame isn't "pump stops, boiler starves, power dies"—it's better to separate insufficient generation, fuel starvation, and night demand as the first read.

I once grabbed Space Exploration info first and thought "pumps need power; if water stops, batteries are the fix." Turned out it was vanilla, so the batteries I added didn't solve it. The real culprit was coal logistics. Early players trip on this swapped assumption a lot.

In expansions and mods, the picture shifts. Space Age regions have different solar efficiency; Space Exploration setups often do require pump power. Those environments have different power self-startup and supply rules. But this guide stays vanilla v2.0. Early-game Nauvis blackouts clear up fast once you master basic steam ratios and know how to read the power screen.

You'll also find old forum posts and legacy version ratios online. This article targets 2.0 vanilla. The real leverage point is always: "In my factory right now, how much kW is being generated, and how much is being asked for?" Close that gap and most early pain vanishes.

Why does early-game power shortage happen? Reading the symptoms

Reading the power screen: fulfilment, production, and consumption

Blackouts are easy to misdiagnose if you're just looking at machines. The power screen's fulfilment ratio, production, and consumption split the problem wide open.

Assembler arms move visibly slower; crafting crawls; research labs slow down. Next, watch whether production is pinned to consumption's ceiling. If generation's line clings to demand's peak, capacity is genuinely exhausted: not enough steam engines, or solar fades at night and accumulators run dry—a classic capacity shortage. But if generation looks like it has headroom yet fulfilment still sags, something else is broken: coal isn't reaching the boilers, an electric pole is unplugged and split the network, or the steam piping is unstable.

A useful reference is the Electric system/ja|Electric system page, which lays out the mechanics clearly.

Accumulators tell a story too. They charge only when there's surplus and discharge only when others fall short. If the graph looks fine during the day but crashes at night, you're short on night-time cover. At 5MJ per unit and 300kW max discharge, they're good for smoothing dips and bridging darkness—not for propping up underlying steam shortages. Taking three screenshots—daytime, night-time, and the moment before the crash—and comparing them reveals patterns: "Does it hold during the day and only drop at night? Is production stuck at ceiling all the time? Does the gauge flail right before collapse?" Each answer points to a different fix.

A useful reference is the Electric system/ja|Electric system wiki page, which lays out the mechanics clearly.

Electric system/ja

wiki.factorio.com

Symptom chains: slow mining → thin coal → dead generation

Early power shortages are nasty because they don't just slow everything equally. They slow the very machines keeping generation alive, which chains the collapse.

Typical pattern: coal power runs on coal mining. When fulfilment drops, your electric miners slow down. Less coal flows. The inserter feeding coal to the boiler starves. The boiler burns less. The steam engines weaken. Generation drops further. A minor gap becomes a near-blackout in seconds.

The trap is that the first sign feels tiny. "Miners are a bit sluggish; the coal line looks thin; inserters sometimes pause." But when that ripple hits the boiler, the whole stack collapses. I've walked into my early base seeing slow research, gone to check, and found the coal line to the power plant shrivelled to half a belt, engines sputtering. The problem wasn't raw generation—it was that fuel delivery runs on the same power the plant itself needs.

The insight is: don't treat each machine in isolation. Miners, belts, inserters, and boilers form a chain. If factory expansion swallows coal supply while the plant's share stays thin, that chain breaks first. "Research got faster, now everything is weird" usually traces back here.

Failure patterns and opening moves

Power shortages look alike but need different first moves, or recovery drags. Early on, sort by the gauge shape into three quick buckets.

Only drops at night but fine during the day? Night-time starvation. You're relying on solar but short on accumulator cover. Daytime generation suffices, but dark hours drain the battery and hit the floor. By vanilla standards, target 25:21 solar-to-accumulator ratio, and for 1MW round-the-clock, about 23.8 solar panels and 20 accumulators. Deviate from that and night-time drops are almost guaranteed.

Constant drop all day and night? Raw capacity shortage. Production line hugs the consumption ceiling and never recovers. Add more generation—steam engines are your tool here, 900kW per unit. If you've added ten 90kW assemblers, you've hit roughly one more engine's load. Ratios make it obvious which expansion tipped the balance.

Graph thrashing—drops, recovers, drops again? Suspect fuel, wiring, or piping instability. Coal trickling in, an electric pole unplugged so part of the grid went dark, or steam pipes clogged and output wavering. A capacity shortage keeps the graph stuck high; an unstable supply bounces wildly.

Your opening move has three steps:

1. Check the power screen: night-only drop, constant drop, or thrashing?
2. Night-only → accumulators and solar ratio; constant → add generation capacity; thrashing → trace coal, poles, and steam pipes.
3. If coal to the plant is thin, stabilise the plant's supply before worrying about the factory.

Lock this logic in place, and "just add more stuff" stops being your instinct. Early power mess is always one of three: capacity, night-time, or delivery glitch. Three numbers on the screen, one fix each.

Steam power basics | the only ratios you need to memorise

Core ratios and max output

Steam is easier to tune if you lock down the numbers early. The main one: boiler : steam engine = 1 : 2. A steam engine outputs 900kW max per unit.

In theory, offshore pumps can supply up to 200 boilers and 400 engines (from the water flow limit), yielding 360MW peak. But real gameplay is cramped and fuel-constrained, so most players stick to neat building blocks—like "20 boilers / 40 engines per pump." That modularity is the practical tool.

The 1-pump block: 20 boilers / 40 engines layout

Don't overthink layout. The standard move is to build the 20/40 module as-is. Line them up: pump → boiler row → engine row, straight and tight. Keep piping in a line; branches clog more easily and mess up your mental model. Steam's stability comes from flow clarity, not just hitting the ratio.

The shape looks roughly like:

Water's edge
[Offshore pump]
      │
  [Boiler][Boiler][Boiler] ... ×20
      │
[Engine][Engine] ... ×40

Coal belt → feeds boiler row

In practice, stack boilers in a line with engines alongside. Feed coal from both sides so the far end doesn't starve. One-sided coal belts leave end boilers hungry, causing output ripples. Early "my numbers look fine but it's choppy" almost always traces to lopsided coal.

Expansion thinking: build in blocks and modules

If you use "1 block per pump (20 boilers / 40 engines)" as your unit, each runs roughly 36MW by rote. When you expand, repeat the full block sideways so piping and coal stay balanced. (This is a bookkeeping convenience, not the theoretical 360MW max—keep them separate.)

クリエイティブマン会員 3DAY最速先行は2/3よりスタート！ ｜ NEWS ｜ SUMMER SONIC 2025 公式サイト

SUMMER SONIC 2025 公式サイト 2025年8月16日(土)・17日(日)の2日間開催！TOKYO／OSAKA

wp.summersonic.com

Fixing power shortages | the recovery sequence

The fast path is: check fuel → manual kick-start → add to ratio → lock supply priority. Blackouts make you itch to build more generation, but usually coal just stopped flowing, a pole is unplugged, or a pipe is crossed. I've rammed hand-coal into boilers in the dark, watched the miners restart, the belt wake, and the whole plant gasp back to life. Work the sequence in order.

Step 1: Check if coal is actually reaching the boilers

First stop: does coal actually get to the power plant? Steam survival depends on fuel line health before it depends on raw unit count. Go to a boiler. Is coal on the belt? Are inserters feeding fuel? Are boiler slots empty? If any of these say "no," even ten more engines won't help.

The trap is starved mining. If coal came from electric drills, blackout stops them, coal dries up, boilers burn out, and the blackout deepens. Chain it: mining → belt → boilers. Miners stopped? Belt empty? Plant's supply chest empty? Pick one; you've found the break.

Also eyeball the power grid wiring while you're there. Boilers run on fuel, not electricity, but if the grid to your factory is split, "fixed generation" still leaves you dark. Are poles disconnected? Is the plant on a separate grid? Spot both now.

Step 2: Manually start boilers to restore power

If the coal belt is dead, hand-feed coal to a few boilers. Kickstart enough steam to light up the grid. Machines reboot. Miners wake. Coal flows back. Your manual shot was the starter fluid.

Priority order: fuel boilers → confirm mining restarts → confirm power reaches the factory. Fidgeting with assemblers won't help if the plant can't self-recover. Watch the sequence: boiler ignites, miners tick, belt creeps, factory glimmers back.

If nothing happens, check steam plumbing. Are pipes connected end-to-end? Blocked? Do wrong bends block flow? Boilers burn but engines don't move = pipe trouble, not fuel. The power screen will show it: generation there but engines loaf.

Step 3: Expand generation to match the ratio

Once fulfilment ticks back to 100%, add steam units by the block, not random pieces. Stacking engines or boilers solo breaks the ratio and guarantees the next collapse. Commit to the unit: if you're working in 20-boiler / 40-engine blocks, append another block. The ratio holds, output scales cleanly.

Make sure new blocks are wired to the main grid and their coal reaches every boiler. Append a coal belt that feeds both sides. A freshly built block left with coal starving at the end won't reach its rated power.

Long-term, add a spare pump unit so generation has cushion. When you later scale mining or smelting, that buffer absorbs the load spike. If night is the weak point instead, accumulators and solar come next. Accumulators: 5MJ per unit, 300kW max discharge. They bridge gaps. Solar: 25 panels : 21 accumulators is the rule. Night demand spikes that? Mix both.

Step 4: Lock coal priority to the power plant

The real anti-crash move is making coal reach boilers before the factory. Early loops happen because coal gets vacuumed into smelting, and the plant starves. Flip the logic: boilers first, leftovers to the factory.

Two ways work. Stick a buffer chest ahead of boilers and load it first. Or use a splitter set to prioritise output toward the boilers. Either way, coal queues at the plant. Smelting slows, no sweat. The plant dies, and your whole loop chokes. Philosophically, the plant is infrastructure, not a side job. Fuel hierarchy reflects that.

This also speeds recovery. Once you hand-restart the boilers and miners wake, coal flows the prioritised path straight back to the plant. The plant stabilises. The factory rebuilds afterward. Make the plant the fuel first priority—that alone is your fastest recovery lever.

Common traps and fixes | layouts that cause blackouts

Coal competition (smelting vs. generation)

The sneakiest pit: coal is being mined but you're still dark. Cause: coal gets absorbed upstream before it reaches the boiler. Smelter expansion happens, stone furnaces multiply, coal gets vacuumed there, and the plant's end starvates. The belt looks alive, so you don't spot that the plant's tail is empty.

Early on I hit this hard. Expanded iron furnaces, coal went there first, plant boilers snuffed one by one. Raw generation units won't fix it. Fuel priority will.

Fix: route coal to the plant as first priority. Splitter with boiler-side set to "priority output"—cheap and effective. Or peel off a dedicated coal line to the plant, with the factory grabbing leftovers. Frame of mind: don't "spare coal for the plant"; instead, "carve out plant coal first."

The difference is huge. Smelt slower = annoying. Plant dark = total reset. Blackout recovery is swallowed by the cost. Design says: generation feeds the grid; the grid feeds the factory.

Pole and pipe gaps (one square kills everything)

Another silent killer: poles don't reach the plant or pipes are cut. Boilers burn, engines idle, factory's fulfilment stays low. Why? Generation's disconnected or steam doesn't flow.

Poles are the culprit most often. The plant sits at a water edge with rough terrain. You add more machines, but one pole got placed just outside the connection range. Or you forgot to chop a tree and the pole didn't land.

Pipes: boiler-to-engine plumbing has a wrong bend, or one segment didn't snap in properly. Boiler steams, engine starves.

Spot them by visually tracing poles from plant to factory grid and walking the pipe route end-to-end. Off by one square? You've found it. The power screen shows it: production looks fine, fulfilment stays low.

Lighting the plant and radar help. Gloomy water-edge plants are hard to scan. Light + radar = gaps pop out.

Accumulator ceiling (300kW/unit limit)

Once you add solar, "panels are stacked but night still drops" comes next. The issue: solar + accumulator ratio is off. During the day, surplus feeds accumulators. Night hits, accumulators drain, but can only push 300kW per unit. A 1MW factory needs 1MW / 0.3MW per unit = 3.33 units minimum just for the discharge rate, even if they're fully charged.

Accumulators hold 5MJ, so night length matters. But instant load spikes can outrun the discharge ceiling even if the tank is fat. Example: night falls, research and assemblers wake, load jumps. Accumulators have juice but can't vent it fast enough. The grid sags.

Solution: don't size accumulators by juice alone. Pair solar and accumulators 25:21, and size the accumulator count to handle the discharge rate your factory actually needs. If factory load spikes hard at night, bulk up accumulator count a bit, or keep steam as a night backup. Steam doesn't have a discharge ceiling—it's flat output.

My instinct: accumulators are a night buffer, not a power plant. Vanilla night needs them, but they're not a battery to run the factory in dark. Use them + solar together, and rest easy. Solo accumulator play = weak.

Transitioning to solar | mid-game stabilisation

Comparing methods: steam vs. solar + accumulators vs. accumulator emergency

Long-term power has three roads. Steam: burn coal nonstop. Solar + accumulators: daytime sun, night cover from stored juice. Accumulators alone: emergency band-aid.

The real payoff is solar + accumulators. A panel outputs 60kW max and never needs fuel once built. No coal logistics headaches means the whole factory gets a lot easier to manage. But panels are useless at night.

The anchor is 25 panels : 21 accumulators. Lock this ratio and day-night balance snaps into place. The daytime surplus charges accumulators; night discharge bridges the dark. I stacked panels first, watched night tank solo, then added the 21 accumulators—suddenly the graph went from jagged to nearly flat. Side-by-side screenshots: "panels only = dropped night graph" vs. "25:21 = steady line" make the difference visceral.

Accumulator-only doesn't cut it long-term. Capacity is 5MJ per unit, but discharge caps at 300kW. Big loads at night blow past that ceiling. Tank has juice, but taps can't flow fast enough. Load spike = blackout despite reserves. Accumulators are for gaps, not backbone power.

Steam is built in and outputs 900kW flat, night or day. Trade-off: fuel logistics overhead. Solar trades space and initial labour for freedom from coal worry. Mid-game the sweet spot is solar as primary, steam as night/spike insurance. That hybrid is the steadiest hull to sail.

Target quantities and day-night cycles

Do the math by needed average load in MW. Per the Power production/ja|wiki, 1MW needs about 23.8 solar panels and 20 accumulators, averaged across a full day cycle.

Want 1.5MW stable? 1.5 × 23.8 ≈ 36 panels; round up to ~54 panels for comfort. 1.5 × 20 ≈ 30 accumulators. Oversizing beats undersizing—demand creeps up as you expand. Tight designs break fast.

Land pressure is real. Steam squeezes near water. Solar sprawls. Plan the solar farm away from where you'll eventually build next.

Also: size by average load, not peaks. A factory that hit 2MW one hour then sits at 1.5MW is different from one cruising at 1.5MW steady. The first can lean on steam for the spike; the second benefits more from solar dominance. Read your base's rhythm.

Why keep steam as backup and how to wire it

Even when solar becomes primary, leave the steam plant breathing. Night dips and load spikes get auto-filled by existing steam. This gives accumulators less work. With steam in the grid, a night shortage or a burst demand momentarily pulls steam online; otherwise it idles. One or two steam engines can do this gracefully.

Your old steam plant is sunk cost turned insurance. 900kW per engine is strong enough for small gaps. Wire it to the main grid; no special plumbing needed. The network auto-routes: solar during the day, accumulator for the dip, steam for the push. You never hand-manage it.

Early solar builders often yank steam fully, then realise: "I hit night and it's close; I wish I had steam." Hybrid running—solar as daily driver, steam as safety net—is slower to set up but way less brittle. For mid-game, that's the play.

Design truth: daylight = solar, night-time and spikes = accumulators + steam. That's stable.

Next steps | expanding factories with power in mind

Expansion rule: more production = more power units too

Here's the trick to scaling without collapse: add production and power units together. Expanding smelting? Add steam blocks at the same tempo. Running a new research line? Grow solar alongside it. Lock that habit in.

The biggest win is smelting expansion. Furnaces eat coal. So do boilers. Build more furnaces alone, and coal gets pulled away, plant starves, grid sags. Expand furnaces + power in sync, and you climb steadily. I call it "more buildings = more power blocks too."

Make steam units your reference scale. 900kW per engine. If you add ten 90kW assemblers, that's ~0.9MW, nearly one engine's worth. Factories feel "small" but their appetite grows fast. Ratios let you feel it coming.

Once stable, layout becomes a main bus design: iron plates, copper plates, electronics down a spine. Sublines branch left and right for local production. Buses scale fast, but they also speed consumption. The rule holds: spine grows + power grows, or the spine kills you.

Research scaling and peak power management

When research ramps, consumption doesn't creep—it jumps. Red and green together is the first big spike. Labs, assemblers, belts, all live at once. So before adding more research, check the power gauge first. If it's tight, slot another steam block before you slot another lab.

I used to do it backwards. "Red runs fine, so green is next; toss in a couple labs." Then mining and assembly all choked—they're upstream of green. New research labs pull from the plate supply, which pulls more power, which shadows the whole line. So fast research needs a fast plate supply, which needs power. Add labs, check the graph. Tight? Generators come first.

Peaks matter more than average. Daytime is fine but noon research setup crashes you? The grid is running at margin. Upgrade generation before labs, and you breeze through. Accumulators alone can't catch sharp demand jumps—the 300kW ceiling won't admit a full spike. But an extra steam engine does instantly. So if mid-game solar is your backbone, keep steam for the bump, and night + peak load share the safety margin.

The research stage is where power shifts from "separate concern" to "design pillar." Reading the graph becomes as normal as watching the belt.

Summary

Power shortages split into supply capacity, fuel starvation, or night-time demand. Early on, hit those three angles and the fix appears. Stick to steam ratios, add by the unit, and you'll scale smooth.

If you black out, restore the plant first: hand-feed coal, relight boilers, fix coal and pole links. After that, restructure coal priority toward the plant, so the same jam doesn't repeat.

As the factory spreads, front-load power growth: expand factory, expand generation. Mid-game pivots to solar + accumulators, with steam lingering as insurance. That hybrid is the steadiest ride to stable late-game, where research and supply both run without hiccups.