Iron bacteria, the source of ore for the furnace Pouring iron bacteria into porous pot The water slowly seeps out of the pot leaving the iron rich paste inside The congealed paste  I'm going to roast the ore to drive off any moisture and impurities to see if it improves the smelt Wood is stacked over a pit in alternating layers  The damp ore is placed on top of the wood pile and a hole is made in the middle to put the fire Fire by friction,  Once you get good with firesticks, it's easier to maintain than a bow drill or pump drill due to fewer parts (rope etc) which is why I don't use the other methods anymore.

The fire is placed in the pit of the ore pile so it burns back down the wood pile roasting the ore Specks of red hot ore fall through the pile into the pit This is done in the hut so the rain doesn't affect the process A few hours later the coals and ore are stirred and separated The ore is grey from the ash but reddish brown when crushed. Charcoal is picked out to be neat but could probably be left in with no ill effect to the smelt Roasted ore and residual wood ash The smelting set up, roasted ore, charcoal, furnace and blower Preheat with wood for an hour Fill to the top with charcoal.

4 single handfuls of ore air blast for about 5 minutes From now on 3 double handfuls of charcoal and 2 single handfuls of ore  After less than an hour and 7 charges of ore the charcoal has burnt down to tuyere level The next day the resulting mass is taken out through the top of the short furnace It is very glassy suggesting a good smelt It's slag with large high carbon iron droplets embedded in it (the grey spheres) Crushing the slag to get the iron out The first nugget of iron.

It can be tested by striking lightly with a stone, slag shatters easily whereas iron doesn't This is larger than typical nuggets from unroasted ore Another large nugget A lot more pieces of iron than from unroasted ore Panning the slag to get more iron still Overall 33g of iron from roasted ore (typically get 15 g from un roasted) Preparing another batch of ore Collecting snail shells to test as a source of lime flux in smelting The idea is that lime reacts with silica in the ore making it more fluid so that the iron comes together more making larger pieces.

Lime should also take the place of iron chemically bound in the slag adding to a greater yield. The shells are crushed to a powder The shell lime is added to the ore prior to the smelt to reduce handling I'm also trying a taller furnace so that all the charges are put in prior to starting. Less interruptions to blowing air will mean a hotter smelt. Furnace is about a meter tall above the air entry  Preheat about an hour again 9 double handfuls charcoal 3 single handfuls ore/shell mix All the materials are put in before blowing the air now so there are no interruptions during the smelt.

7 charges in total The smelt goes uninterrupted for about the same time as usual The product needs to be taken from the door in the base of the furnace now as it is too tall to be taken from the top The mass is jagged and not glassy like the last smelt No iron is evident prior to crushing, this is not a good sign A tiny iron prill This is all that came from the smelt, adding shells was a bad idea Compared to last smelt, adding shells greatly reduces efficiency. This is probably due to the ore already being basic. Increasing the PH made the slag less viscous  Discarding the lime ruined slag.

I'll try re-smelting the original slag (no lime) and see if any more iron comes out Re-forming the door Hour preheat again 9 double handfuls of charcoal 3 single handfuls of slag then fill the rest up with alternating layers: 3 double handfuls of charcoal and 3 single handfuls of slag 9 Charges in total The product is better than the previous smelt Iron is immediately visible in the slag (grey spheres).

A good yield considering this was from waste slag from old smelts Roasted ore smelt on right: 33 g. Recycled slag smelt on left: about 20g All the iron I currently have at the moment from various experiments