‹ Back

Second firing of the earth kiln

 

An unusually dry - and it turned out beautiful - day in an otherwise rather damp month, gave the opportunity for a second firing of the 'medieval' earth kiln made last year. The previous day we had dodged rain showers to pack the kiln, using pots made both last year and this, in the hope that the weather would live up to the promise of the forecast. We were not disappointed. 

 

The set up of the kiln was largely unchanged from last year. We had discussed the option of blocking up the middle archway under the firebars in the base of the ware chamber, to force flame upwards even when gusts of wind blew into one of the fireboxes, but decided against it as it probably also has a role in balancing the input from the two fires. The only real change was that cob bars were used to raise the ware chamber floor slightly, and create a baffle to split and disperse the flame path from the top of each firebox, which we hoped would ensure the pots nearest the fireboxes would experience more even heat than last year.

  

Packing the kiln was made somewhat easier by having fewer types of pottery than last time - mainly cooking pots and just a few jugs - but not helped by doing it mainly in the dark, and the fragility of the rims of the cooking pots!  As previously, we used a layer of broken fired pottery as the top layer of the pack to act as a heat retainer, but tried to set this with bigger pieces to achieve a better airflow than last time.

We had let modest fires burn for a few hours in both fireboxes earlier in the day so the kiln was already warm when packed. Some of that heat was still present at 07:00 the next morning, despite the thick mist and heavy dew, when the fires were lit for the actual firing. Probably for this reason, in combination with a deliberately more open packing of the ware, the initial small fires - intended to keep the kiln around 100-200°C - burned steadily and were easy to control. This was helped by the fairly still conditions, until the mist cleared at about 08:30 and a modest south-westerly breeze settled in for the day.

For the next 3h or so we let the temperature climb very slowly, to about 200°C, then gradually raised the stoking rate to give a steady climb of about 50°C per hour until by 16:00 the temperature was about 800°C. One of the issues we had encountered with the first firing was the kiln going into periods of heavy reduction (lack of oxygen in the kiln), and this caused us some problems with some of the clays we were using. This time we were more careful with the stoking and managing the airflow in the fireboxes, in addition to having had a somewhat looser pack in the kiln, and signs of reduction were occasional and short-lived. As the temperature slowly climbed and the flames increasingly reached further up into the ware chamber, the effect of the new firebars on the flame path could (just) be seen.

Although the temperature rise was fairly steady overall, the pyrometer, which was sensitive to the changes in flame passage through the ware chamber highlighted the fluctuations resulting from both the stoking cycle, and our moments of inattention: the light wood we were now using burned out very quickly as soon as it was in contact with the deep hot bed of charcoal. To  keep a clear airflow, we found that using a metal rod with hooked end it was possible to rake the bottom layers of the charcoal back towards the stokehole, lowering the overal height of the bed but leaving the hot burning layer on top reasonably undisturbed. This helped keep the fires burning more evenly. 

 

By about 18:00 the temperature was about 1000°C, and we decided to extent the exit flue upwards to encourage draw through the ware chamber, as the breeze was dropping with the lowering sun. A simple but effective construction of cob blocks and bricks, had the desired effect and over the next couple of hours the kiln held a steady orange heat, fluctuating around about 1040°C. The heightened exit flue was an interesting experiment we had not tried previously, but it was a simple, and effective technique, which could easily have been used by medieval potter too. 

 

Pyrometric cone 05, our target, went down about half-past eight, and we gently eased off fuelling the fires, and let them  subside to a deep bed of red hot charcoal, the chamber temperature dropping steadily as the flames died (though the fabric of the kiln and the pots would have fallen much less quickly). Once the flames had died down, both stoke holes were roughly sealed with bricks and clay cob, and the bricks around the exit flue removed so that could also be covered. The firing had taken about eleven and a half hours.

By the middle of the next morning the chamber temperature was about 200°C, dropping to about 80°C the morning after that, at which point we opened up a couple of bricks in the stoke hole entrances, and lifted the flue cover by an inch or so, to aid the last stage of cooling. By early afternoon the kiln was about 40°C, and we were able to open up properly.

There were a couple of breakages, one where there was a stone buried in the wall of the pot and another where a split developed in the wall of a small pot, but otherwise all the pots emerged intact and generally well fired. There was evidence of only light reduction in places, so it looked like our care in controlling the stoke rate and airflow had been successful. The two main issues we had were that the clays we were using were clearly just starting to soften at the temperatures reached at least in parts of the kiln, so some of the cooking pots were somewhat less circular than when they went in, and where pots had been resting on the rims of others the rims were sometimes dented. This suggests that either some parts of the kiln got hotter than the cones indicated, or that the clays we were using simply need firing to a slightly lower temperature (or shorter time at the upper temperatures). The other issue was with the glazed pots. In several cases the glaze on parts of the pots (usually one side) had overfired and blistered. In most of these cases on the other side of the pot the glaze had fired more or less as expected - again an indication of uneven temperature exposure.

Cooking pot - example of a well fired piece

 

Cooking pot with chipped rim (from loading) and distortion where clay has softened

 

Medieval flower pot - well fired

 

Cooking pot damaged by a stone in the pot wall

 

Jug - well fired except for slight overfiring of the glaze on one side

 

There doesn't seem to be a clear pattern to where in the kiln the pots that have fired well, or poorly, were located, so it seems that the hot spots may just be locations where there was a more direct flame path through the pots. The main conclusions for next time are that we may need to hold the kiln at a slightly lower temperature for these clays, and that glazed pieces may need to be placed in areas less subject to direct flame. More care in loading the kiln - avoiding many of the chipped rims - would also be good! Overall, however, a successful firing and definite progress in understanding how this kiln, and the clays we are using, work together.