For a more detailed Field Report please see the Bulletin of the NCEFG Vol. 3.
Diagram of our kiln - Stanley IIb.
Stanley IIb: A Survivor
 We were delighted to discover that despite being completely under water several times during the long winter (there were even resident fish at one point), LINK Stanley IIb came through the rainy season with flying colors. The kiln displayed complete structural integrity in all areas. (Though it must be admitted the firebox smelled rather badly of dead fish.)
We performed minor repairs to the interior of the ware chamber in which cracks remaining from the previous season had propagated during the winter.
So, in 2005 the NCEFG was able to use the same kiln as in 2004, Stanley IIb, a single-firemouth updraught design with raised ware chamber and discrete firebox.
 As a quick reminder, Stanley IIb's ware chamber was made up of a solid clay interior (fired to ceramic during in the first firing of Stanley IIa in 2004) surrounded by a thick outer shell of coursed stone in a clay and straw matrix.
The firebox was made up of firebrick mortared and covered with the clay and stray matrix of the ware chamber. This was a deviation of the NCEFG's charter to use all natural materials. In 2005 we discovered (to our horror) when not one, but two people stood on top of the firebox roof, that it was still quite load bearing. We continued to use fragments of old kiln shelves to make up the ware chamber floor, and to lid the top.
Thus before being put back into action in 2005, Stanley IIb's ware chamber had been fired three times and patched twice, its firebox had been fired twice, and lower underground portions had been fired six times over three years.
 We did make one major modification. With our first disastrous firing of Stanley IIa in 2004 we'd decided not to include at thermocouple port in Stanley IIb. With all the damage from that first firing we did not wish to further risk the integrity of the ware chamber. However, with such a successful winter survival and the acquisition of a third thermocouple we drilled a small hole on the west side of the ware chamber. This may, or may not, have had significant effect on the interior hot spots and heat circulation in the kiln interior during our 2005 firings.
Materials
Clay
 Original construction: The interior of the ware chamber was made entirely of clay and straw briquettes with no stone except as support over the firebox throat. We used recycled shop clay, Pueblo White, mixed with a local yellow earth clay (an upper sedimentary deposit from near-by Peacock Gap, Marin County, California), and sandy dirt (on site deposits), and straw (as opener). The exterior of the ware chamber was made up of courses of Sonoma fieldstone mortared together with a mixture of kaolinite clay (quarried from the Sierra Nevada mountain range in Amador County), the local yellow clay from Peacock Gap, sandy dirt, and straw.
Repairs: The mixture used to repair Stanley IIb in 2005 was comprised of shop clay from the 2005 season's pottery construction (both white and orange clays with sand, talc, and grog tempering), mixed with equal parts fine grain river sand and fresh grass clippings. The grass proved an excellent alternative to straw. It dried at the same rate as the wet clay, was less resistant to shrinkage, and was small enough to fit inside cracks in the ware chamber. All this and it still served the purpose of an organic inclusion: to prevent too much shrinkage and crack propagation.
Stone
The same stone was used this year as in 2004.
Fuel
 We used the same wood fuel in 2005 as we did in 2003 and 2004: woodshop kindling and fast-burning high-heat eucalyptus logs in three sizes. We found the weight method of tracking wood successful in 2004. Directly after cutting we divided the fuel into 9.2kg (20lbs) piles. As always with experimental firings, wood chopping took a considerable amount of time. Luckily, we were blessed with some serious wood chopping mavens this year.
Brick Making
 In 2005 we decided it was time to make our own bricks to replace the modern firebrick we'd resorted to in LINK 2004. We mixed together two buckets of shop clay, one bucket of crushed kaolinite (the last of our Sierra Nevada stores from 2003) and one bucket of the Peacock Gap yellow earth clay. We formed approximately thirty bricks by combining the above mixture around an internal body of wet straw.
At first we tried forming the bricks by hand, but soon enough participants had constructed a wooden form out of our shop kindling, which could make three bricks at a time.
In the end, all of our brick making was for naught. Even after a week of drying (in the living room next to the fireplace - much to A. Vernon's horror) they were still far too damp to fire in the kiln. They were left in the studio, where they would definitely be ready by 2006.
Pottery
 To make the pottery we used a wide range of different wet shop clays. We had two low-fire, three medium-fire earthenware, and one high-fire wide-range stoneware at our disposal. Our low-fired clays were comprised of: a talc-tempered soft grey clay, and a no-talc fine grey clay. Our three medium-fire clays were: a beautiful dark red grog-tempered southwestern clay; a grog-tempered orange clay; and a sand-tempered orange clay. The high-fire clay was white in color with a very high grog and sand content. To make the pots we used the on site pottery studio.
The pots were constructed using the coil, slab, mold, and wheel throwing methods. Most pots had at less than a week to dry, although several had been sitting around for months. We should have learned our lesson in 2003, for as with that season we would loose many pots in our bisque firing due to wetness and bad spalling.
Firing Stanley IIb: Bisque
The day of our first firing in 2005 was quite foggy and brisk. The quantity of fog meant very little wind. These were not the most ideal of firing conditions.
 We had enough pottery to load the kiln full to the top, and were ready to commence firing at 10:30 AM on July 2, 2005. Since they had worked so well in 2003 and 2004 we used the same two types of temperature measuring devices: type K thermocouples and Orton pyrometric cones.
We used four banks of Orton pyrometric large size cones in rows of six each. Unlike previous seasons each bank did not contain exactly the same range of cones. Since we knew most of our heat was concentrated near the bottom, we place a high cone bank at this level, medium in the center, and low at the top, thus covering a temperature range of 985°C to 1160°C.
This year we had three sets of thermocouples plus attached analog meters: one placed in our newly drilled thermocouple port, and two positioned to tilt down into the ware chamber from the top edge of the kiln. These gave us a general idea of the internal temperatures of the kiln as firing progressed, but (as we would learn later from the cones) were cooler than actuality.
We took the initial temperature up slowly so that after about five hours we'd reached 300°C, with the first hour kindling, the rest of the time small sized eucalyptus, amounting to 30kg (approx. 66lbs) at the end of five hours. For most of this time there was little or no smoke. We referenced a 2004 firing graph (from our supremely successful bisque) and attempted to keep our firing on a similar path. Although we were going even slower than in previous years we heard the first ominous "poof" of pot explosion at about 1:00 PM (2.5 hours into the firing). We instantly stopped stoking and slowed firing but continued to hear explosions for the next hour or so.
At 2:30 PM smoke became visible out of the thermocouple hole, and a second boom emanated from the kiln, Though we corked around the thermocouple as much as possible, it would continue to smoke and even steam for the remainder of the firing. Smoke also began escaping various other cracks in the kiln. We used leftover patching material to plug them up, right then and there, during firing.
At 4:30 PM and 400°C dark smoke began to emanate from the top of the kiln. We continued to stoke for the next several hours. By 8:00 PM flames about two inches high appeared above the lip of the ware chamber alternating with gouts of black smoke as the kiln burned away carbon deposits. Thermocouple B ceased to operate properly (our problem-child thermocouple from 2004), but the other two confirmed ceramic temperature. This "black smoke marker" was the same as in previous bisque firings.
 In ancient times (with no thermocouples to rely on) it's easy to believe that it was not just the color of the pottery but the behavior of the kiln that indicated when ceramic had been reached. In our bisque firings a white or grey smoke and steam (accompanied by the "poof") indicted boiling point and the first major stage of firing (300-400°C), while black smoke and flame out the top of the kiln (600-700°C) indicated ceramic.
We continued to push Stanley IIb and reached our highest thermocouple reading, 1075°C, at 10:00 PM -11.5 hours into the firing, after which the kiln stalled and all attempts to raise the temperature further proved ineffective. We capped over the top of the kiln and shut Stanley IIb down for the night.
Results of the Bisque Firing
The bisque firing took 12 hours to complete. We used 180kg (about 397lbs) of wood, which was less than our 2004 bisque firing which used 202kg (440lbs). The highest temperature measured by a thermocouple (Thermocouple C at the port) was 1075°C, while in 2004 it was 750°C. (This difference can be accounted for by the fact that in 2004 there was no thermocouple port so all measurements had to be taken from the lip, a decidedly cooler location.)
 However comparisons of the cones of the two seasons showed that in 2004 we actually reached higher temperatures than in 2005. Also in 2004 three out of four cone banks melted to at least 02 (approximately 1100°C) a nicely consistent internal temperature, while in 2005 we saw nothing of the kind. Two banks reached 05, but the two higher placed banks did not melt at all. Since the melting of cones is dependant upon length of exposure to a specific temperature as well as reaching that temperature initially, we must assume that in 2004 we were able to better maintain temperatures than in 2005. (Could this be the thermocouple port's fault? We don't know.).
 Perhaps we were cocky from our successes in 2004, but this bisque firing had a much lower survival rate. As we unloaded we encountered some serious carnage. The floor of the ware chamber was littered with pottery fragments and shards. We'd begun with 58 pot in the kiln, only 29 would have been sellable (by an ancient potter), which meant a 50% survival rate. Most of those lost were the large saggars. Because of their size there were both less dry and were being used to hold up shelving and shelter other pots, so that when they spalled or exploded they inflicted the maximum amount of damage.
 Since our main difference in pot survival between 2004 (95%) and 2005 (50%) was damp weather and damp pottery, we attributed most of our poor survival rate to wetness. In the future we intend to divide the NCEFG season into two parts: a two week pot construction run, a months drying time, and then a two week firing run. This should eliminate this problem entirely.
The kiln itself survived this firing well. Only a few minor repairs were needed to prepare Stanley IIb for a glaze firing.
 
Pots before and after firing (bucket of fragments in foreground.)
Firing Stanley IIb: Glaze
In general our biggest problem with loading and firing glazed pottery is that pots can not touch other pots. This means that in general we can glaze only about half of the bisqued pots. It also means that we required a considerable number of saggars, shelves, and stacking units. We'd been saving these up from previous years but had banked on having several new saggars this year. Unfortunately it was mainly these units that had exploded in the first firing.
 Despite this we managed to fit 36 glazed pots into the kiln (using unglazed pottery from previous seasons as well as our 29 survivors). We placed four cone banks carefully at various point in the kiln. Loading took three times as long as loading for bisque.
As always with glazes instead of pushing our kiln to new heights, we had to think in terms of reaching a specific temperature (between cone 06 and cone 04, approximately 995-1060°C) and hold it there for several hours.
 In 2004 we'd had success raising the temperature more rapidly than we felt was safe and we did that again in 2005. In four hours Thermocouple A was at 900°C (the port) and B was at 785°C (the ware chamber rim). We maintained temperatures in Thermocouple A's 900 range for the next 2.5 hours, ending the firing in late afternoon, after just over 6 hours.
Results of Glaze Firing
 We used approximately 160kg (353lbs) of wood this was less than in 2004 when we'd used 197kg (428lbs) over six hours. The highest temperature measured by the thermocouple was 985°C (higher than in 2004, but this is probably because we access to a thermocouple port in 2005 and not in 2004).
 The highest temperature measured by the pyrometric cones was 1 (approximately 1154°C) and the lowest was 06 (995°C). There was an obvious discrepancy between the bottom and top of the kiln, indicating quite a range inside the ware chamber.
 Unfortunately as we unloaded the kiln we realized that we had pushed temperatures too high. Consequently, the resulting glazes were quite bizarre looking, displaying all manner of over-fire damage including: pin-holes, crazing, crawling, and bubbling.
 
Two spectacular examples of over fire. First image shows ubbles and white color developed by a clear unleaded glaze. Second is a Typical example of crawling, when glaze reaches a temperature so high it can no longer bond properly to the pot.
The wide variation of temperatures and atmospheric conditions inside the kiln also meant that a single glaze could (and did) turn out a spectrum of coloration. For example our blues and greens (whose main color pigmentation comes from copper) turned or developed a metallic patina of coppery red.
A subtle blue green glaze is turned by our enthusiasm into a wide range of colors.
  The kiln itself survived this firing with the expected array of cracks, but no major damage. Consequently we shut Stanley IIb down for yet another season, left to weather the winter months once more.
Not with a Bang but with a Whimper...
 It must be concluded that the 2005 produced less successful pottery than 2004 but was still quite interesting from a scientific standpoint. We proved that a kiln could safely weather a long rainy season and fire properly with only minor repairs. We also showed that those repairs could take the form of clay patches onto and thrice fired ceramic body and still adhere without spalling.
However we did learn that it would be better in the future not to push the kiln towards ever higher temperatures, but instead to go for length of time at a consistent temperature of 750-800°C. As we are attempting to replicate ancient methods of firing we should begin using visual clues, rather then relying on thermocouple readings and graphs, to predict our stopping points. For example, the point at which black smoke stops emanating from the ware chamber and flames begin to appear above the lip is a good indicator that ceramic has been reached and no higher temperature is needed.
For glaze firings we must keep a close eye on those pots that can be seen over the lip of the ware chamber to watch for bubbling. Also we must conclude that saggars are a good idea, the more the better. By far the most attractive glazed pots in 2005, although still decidedly over-fired, were those that had been saggared.
We must also return to our discovery of 2003, that wet pots are the most prevalent danger to survival rates. Our pottery must be allowed greater drying time than one week, especially if the weather is uncooperative.
All these methods should result in a higher survival rate. In 2006 - we better ourselves!
For more scientific conclusions and analysis, as well as additional details, please see the 2005 Field Report in the Bulletin of the Northern California Experimental Firing Group Vol. 3.
|
