All guides

Start here

Fundamentals

Materials

Technology

Knowledge

Fundamentals: Fire, Fuel, and Reaching High Temperatures

Difficulty: moderate, and it is the skill everything else waits on. Prerequisites: an axe, a shovel, clay, and hide or cloth for bellows.

Two hinged-board leather bellows on a trestle, rocked by one beam, blow through a clay Y with a tell-tale flap and a down-angled tuyère into a half-sectioned clay shaft furnace, where the charcoal around the tuyère mouth glows with the air that reaches it.

Interactive figure, Fig 4 · Twin Bellows Blast: set the stroke rate from 4 to 20 a minute, rock the beam by hand, and switch between one bellows and two worked alternately to see the tell-tale flap and the bed's glow pulse to nothing between strokes or stay steady at the same mean air. A code-drawn explanatory model built from this guide's numbers, for orientation only; it is not an equipment drawing or a procedure.

Why it matters

Almost every guide in this repo depends on hitting a specific temperature and holding it. This is the single most limiting skill for a time traveler, and the one where knowledge helps most — ancient people had the same wood you do, and mostly didn't know how to make it work harder.

Get this page right and you unlock pottery, lime, glass, copper, iron, and silver. Get it wrong and none of them are possible at all.

The temperature ladder

Memorize roughly where things happen. Everything else is engineering to get there. This table is the repository's reference; where a guide quotes a different figure, it's describing a specific stage of its own process.

Setup Temp What it unlocks
Open wood fire 600–900°C Cooking, pottery bisque, charcoal burn, gypsum plaster
Simple updraft kiln, wood 900–1000°C Earthenware pottery, roasting ore
Lime kiln, held for 8–24 h 900–1000°C Quicklime — the long soak is the hard part, not the peak
Charcoal + natural draft 1000–1100°C Melting lead (327°C), tin (232°C), antimony (631°C), silver (962°C)
Charcoal + bellows 1100–1300°C Melting and casting copper (1085°C) and bronze, smelting copper ore, glass melting and fining, bloomery iron, cupelling silver
Charcoal + strong bellows 1300–1400°C Stoneware and porcelain, difficult smelts, melting cast iron comfortably
Charcoal + strong bellows, small sealed crucible 1500–1600°C Crucible (wootz) steel — the hardest thing on this list

Two things people get wrong here. First, cast iron melts at only ~1150°C, not 1500°C — that's the whole reason the Chinese blast-furnace tradition worked (see iron and steel); it's pure iron at 1538°C that's out of reach. Second, the number that matters for most processes is not the peak but how long you can hold it. A furnace that touches 1300°C for five minutes will not fine a glass melt or burn limestone through.

Within a fuel's correct combustion range, controlled airflow and lower heat loss can raise useful process temperature. Too little air leaves fuel unburned; too much can cool the charge, waste fuel, or create the wrong atmosphere.

Step 0: Get a flame at all

The rest of this page assumes you already have fire. If you arrive with nothing, that assumption is wrong, and this is the first problem to solve. It is also the one where a modern reader is most likely to be overconfident, because striking a match has never been the hard part.

Carry the fire rather than restarting it. Every method below is slow and unreliable in bad conditions. The historically normal practice was to keep a fire alive continuously and to borrow embers from a neighbour, and a banked fire — fuel raked into a mound and covered with ash so it smoulders overnight — is far easier than lighting a new one. A fire pot (embers in a perforated clay or horn container packed with punk wood or moss) lets you carry live fire while travelling. Treat re-ignition as the fallback, not the routine.

Tinder, kindling, fuel — get this order right

Most failures are fuel-structure failures rather than ignition failures. You need three grades ready before you make a spark:

  1. Tinder — takes a spark or a coal and turns it into flame. It must be bone dry and finely divided: cramp ball fungus, dried Fomes fomentarius amadou, birch bark (which lights even damp, because of its oils), cattail or thistle down, teased-out inner bark, dry grass, charred cloth.
  2. Kindling — pencil-thick dry twigs, split softwood, resinous fatwood.
  3. Fuel — progressively larger split wood.

Char cloth is the single biggest improvement available and it needs no special material. Put scraps of plant-fibre cloth in a nearly sealed tin or clay pot with one small vent, heat it in a fire until the smoke from the vent stops, then seal the vent and let it cool. The result catches a faint spark that would die on anything else. Making char cloth requires an existing fire — which is exactly why you make a batch every time you have one.

Methods, roughly easiest to hardest

Chemical ignition, and why matches are later than you think

Friction matches are 1820s–1830s and depend on chemistry this repository only partly supports. Early phosphorus matches also caused "phossy jaw" — disfiguring necrosis of the jaw among match workers exposed to white phosphorus — which is a case worth remembering when weighing a convenient product against what making it does to the people who make it.

Fire safety is not a footnote here. An uncontrolled fire is one of the most destructive events a pre-modern settlement can suffer, and whole cities burned repeatedly. Clear a margin down to bare soil, keep fires small and contained, never leave one unattended near thatch or timber, keep water or sand within reach, and site workshops that use fire away from dwellings and downwind. The carbon-monoxide warnings later on this page apply from the very first fire you light indoors.

Step 1: Make charcoal

Wood contains water, tar, and volatiles that consume energy and make combustion harder to control. Charcoal is carbon-rich and can provide a hot, low-ash, repeatable furnace fuel, but it does not burn cleanly: it produces lethal carbon monoxide and can produce particulates and other combustion products. Many historical metallurgical and glass processes used charcoal, while suitable coal, coke, wood-fired kilns, and other systems were also possible. OSHA's carbon-monoxide guidance explains why odor, flame behavior, and apparently adequate ventilation are not exposure tests.

Pit method (simple historical method; not suitable everywhere)

  1. Dig a pit roughly 1 m across and 0.6 m deep, in soil that isn't waterlogged.
  2. Fill it with dry hardwood split to roughly wrist thickness, packed tightly and stacked in one direction. Dense hardwood (oak, beech, hornbeam) makes hard charcoal that survives forced air; softwood makes crumbly charcoal that collapses and blocks airflow.
  3. Light it from the top and let it burn openly until the whole surface is flaming and the wood is charring — usually 30–60 minutes.
  4. Cover with green leaves/grass, then a layer of soil, leaving a few small vents. The goal now is pyrolysis, not combustion: heat drives off volatiles with almost no oxygen present.
  5. Watch the smoke. Thick white smoke = water and tar still coming off. When the smoke thins to a faint blue haze, the wood is charred through.
  6. Seal all vents completely with soil and leave it 12–24 hours to cool. Opening it hot means it bursts into flame and you lose the batch.

Yield: roughly 20–25% of the dry wood weight — so you throw away three quarters of the mass to get a quarter that burns properly. (By volume the loss looks smaller, around half, because charcoal is far less dense than wood. Weight is the honest measure and the one to plan with.)

Good charcoal is black through the middle when snapped, rings when dropped, and leaves little ash. Brown centers mean you undercooked it — those pieces still contain water and will not reach high temperatures.

Mound (kiln) method — better yield, needs practice

Same principle, above ground: stack wood into a dome around a central chimney shaft, cover with turf and earth, light through the chimney, then control the burn with vents around the base. Yield ~30% by weight. This is what medieval colliers used and it is worth graduating to once you need charcoal in volume.

Charcoal maths, and the constraint nobody plans for

Smelting a kilogram of metal typically eats several kilograms of charcoal, made from 15–20 kg of wood. Charcoal demand, not ore, is what limits pre-industrial metal production.

This is a renewable constraint, not a fixed one, and the difference is the whole game. A furnace fed by clear-felling exhausts its landscape in a generation and takes the community's timber, fuel, and soil with it. The same furnace fed by a managed coppice runs indefinitely. Before you scale up any metal or glass production, read woodland and fuel supply — it tells you how many hectares a given furnace actually needs and how to cut wood so it grows back forever.

Site the furnace near the woodland, not near the mine. Wood is bulky and ore is dense; move the ore.

Wood tar and pitch — the same fire, a second product

Pyrolysis drives tar out of the wood. Collect it instead of losing it and you get the pre-industrial waterproofing and preservative:

Historically, a sloping impermeable floor and drain collected condensate from a sealed charcoal or tar kiln. That description is not a safe open-workshop recipe. Hot tar and its vapor can burn, ignite, and expose workers to complex irritant and carcinogenic mixtures. Collection and concentration require an engineered closed or source-captured system, indirect controlled heat, remote handling, fire containment, exclusion of bystanders, and hazardous-waste controls. Do not boil tar in an open pot or over a direct flame. NIOSH treats coal-tar-pitch volatiles as an occupational carcinogen; wood-tar composition differs, but unknown tar smoke and condensate likewise need process-specific characterization and control rather than sensory judgment.

Properly produced tar can be concentrated to pitch. Between them they seal boats, preserve rope and timber, line pottery, and make the pitch lap for grinding lenses. Birch bark gives a superb waterproof tar and is the oldest known adhesive.

Other fuels

Step 2: Force air into the fire

Natural draft gets you to about 1100°C in a good chimney. Beyond that you must push air.

How much air? This is the number nobody gives you, so: a small shaft furnace of 25–30 cm internal bore needs roughly 300–500 litres of air per minute to run a bloomery smelt. That is two people working alternate bellows steadily, and it is genuinely hard physical labor for six to eight hours. A small crucible furnace needs perhaps 100–200 L/min. If your furnace won't reach temperature, insufficient air is the most likely cause, followed by heat loss through uninsulated walls.

Bellows. The classic pre-industrial device: a leather bag or two hinged wooden boards joined by leather, with a one-way flap valve for intake and a clay or metal nozzle (tuyère) for output. A pair of boards about 60 × 40 cm hinged at one end, with a stroke of 30 cm, moves roughly 30–40 litres per stroke — so one bellows at 12 strokes a minute gives you around 400 L/min. Size yours from that.

Use two bellows worked alternately — a pulsing single bellows lets the fire cool between strokes. In many cultures a boy on the bellows was a full-time job; budget the labor, and put the bellows on a waterwheel as soon as you possibly can. That single change is what made the European blast furnace possible.

If you have no leather (and note the circularity — leather needs lime, lime needs a kiln): a hollow log with a plunger wrapped in cloth works, as do the paired-drum bellows of West African and Southeast Asian smiths (two clay pots with skin diaphragms, pumped by hand), and the Chinese double-acting box bellows, which delivers air on both the push and the pull stroke and is the best hand design there is.

Nozzle placement is critical. The tuyère should enter low, angled slightly downward into the fuel bed, positioned so the blast passes through burning charcoal rather than over the top of it. Air over the top just cools the charge. Protect the nozzle tip with clay — it will otherwise melt.

Step 3: Build the furnace

Clay

Finding, testing, and processing clay is covered properly in clay and earth building — the ribbon, shine, and jar tests, and the slake-and-settle refining loop. Read that page first.

For furnace work specifically, you want pale, high-alumina, low-iron fireclay (often found under coal seams) rather than the iron-rich red clay that's fine for bricks. Iron and lime act as fluxes: they'll make your furnace lining melt at exactly the temperature you're trying to reach.

Temper heavily with grog — crushed previously-fired clay, at 20–40% for furnace walls and 30–40% for crucibles, which take the worst punishment. Grog has already done all the shrinking it will ever do, so it acts as a skeleton. Every broken pot goes in the grog pile.

Furnace geometry

Insulate. A furnace packed with earth, ash, or a double clay wall reaches several hundred degrees higher than a bare one with the same fuel and air. This is free temperature and the most commonly missed trick. Wood ash between two clay shells is an excellent, free insulator.

Dry it slowly. Air-dry a new furnace for days, then fire it gently for a full day before working it hard. A furnace fired fast while damp cracks itself apart, and that is how most first attempts end.

Step 4: Read the temperature without a thermometer

Judge by color of the glowing interior, viewed in shade (daylight makes it look cooler than it is):

Color Approx. temp
Faint red, barely visible in the dark 500–600°C
Dull cherry red 700°C
Bright cherry / light red 800–900°C
Orange 1000–1100°C
Yellow 1100–1250°C
White, painful to look at 1300°C+

Published melting points are useful calibration references, but casually melting scraps in a furnace mouth is not a safe thermometer. Lead and antimony can create toxic fume, dust, and molten-metal exposure; salt can spit or contaminate work; and alloys or impure scraps do not melt at the pure-substance value. Prefer a calibrated thermocouple or commercially characterized pyrometric cones. Treat the following as reference values, not instructions to expose a marker. OSHA's lead guidance requires exposure and hygiene controls beyond merely working outdoors.

Marker Melts at
Lead 327°C
Antimony 631°C
Common salt 801°C
Silver 962°C
Copper 1085°C
Cast iron ~1150°C

For repeatable work, use characterized pyrometric cones appropriate to the heating rate and atmosphere. Homemade clay/flux pyramids can be useful comparative batch indicators, but a one-point metal observation does not turn them into a traceable temperature scale. Compare them against calibrated instrumentation or certified cones over the intended firing cycle, record the recipe and conditions, and recheck new material batches. See mathematics and record-keeping — this is exactly the kind of thing to write down rather than trust to memory.

Oxidizing vs reducing — the control everyone misses

At similar indicated temperatures, changing air-to-fuel ratio can drive very different oxidation, reduction, color, and combustion outcomes. Atmosphere is a process variable to measure and control, not a substitute for a complete recipe.

Control it by the depth of the fuel bed above the tuyère and by choking the air supply. In a shaft furnace, keep the stack full of charcoal so the blast has fuel to burn through: air entering low and passing up through a deep bed of charcoal becomes reducing by the time it reaches the charge. Run the stack low and you re-oxidize your metal and lose the smelt.

Safety, honestly

Failure modes

Symptom Cause Fix
Charcoal is brown in the middle Burn stopped too early Wait for the smoke to thin to blue haze before sealing
Charcoal crumbles to dust and blocks airflow Softwood, or over-burnt Dense hardwood; seal as soon as the smoke turns
Charcoal batch burns away to ash Vents not sealed, or opened hot Seal completely; leave 12–24 h to cool
Furnace cracks apart on first firing Fired fast while damp, or untempered clay Air-dry for days, fire gently for a day; 20–40% grog
Furnace lining slumps and melts Iron-rich or limey clay acting as flux Pale high-alumina fireclay, heavily grogged
Furnace never reaches temperature Too little air, then too much heat loss More/bigger bellows; insulate the outside; check the tuyère isn't blocked
Tuyère tip melts away Unprotected nozzle in the hot zone Clay sleeve, replaced each smelt
Blast seems strong but the charge stays cold Air passing over the fuel, not through it Lower the tuyère; angle it down; keep the stack full
Metal oxidizes instead of smelting Oxidizing atmosphere — too much air, too little fuel Deeper charcoal bed, choke the air
Temperature swings wildly Single bellows, pulsing Two bellows worked alternately, or a box bellows
Enormous fuel use, woodland shrinking Unmanaged cutting Coppice — see woodland and fuel

Sources and limits

The temperature ladder, charcoal yields, example furnace dimensions, and airflow figures above are legacy order-of-magnitude planning estimates. This review did not validate them as design specifications across fuels, ores, refractory materials, instruments, altitude, scale, or operators. Measure the real process, record uncertainty, and add a guide-level authoritative source before treating a number as a transferable operating limit. The repository-wide source and safety boundary applies to all high-temperature work.

Era notes

Controlled fire is older than our species. Charcoal production, shaft furnaces, tuyères, and bellows are ancient, and charcoal was important to many historical metallurgical traditions. Archaeological methods and fuels varied by place and period, so the pit and mound descriptions here are examples rather than a single Bronze-to-medieval design.

Your knowledge advantages: understanding oxidizing versus reducing as a deliberate control rather than luck; insulation as efficient heat retention; knowing that cast iron melts far below pure iron; calibrated temperature instrumentation and characterized cones instead of guessing by eye; putting the bellows on water power; and above all coppicing your fuel supply so the furnace doesn't eat the forest that feeds it.

Where to go next