anagama kilns
Wysing kiln plans
Loughborough kiln plans
Chemistry of fly ash

Diverse atmospheres

This text discusses the relationship between kiln design and kiln atmosphere

There are very few other firing environments that are capable of developing the types of atmospheres and therefore clay body colour responses that those generated in a single chamber woodfired cross draft kiln. All fuel-burning kilns have some degree of control over the fluctuation between an oxidised atmosphere (one containing free oxygen), and a reducing atmosphere (one starved of oxygen). However, these are usually controlled by the adjustment of dampers. Once these setting have been instigated it takes another deliberate re-setting to change the atmosphere again. Within the chamber of an anagama this balance between oxidising and reducing is in a constant state of flux. To a certain degree, depending on the design of kiln, the above controls can also exist. However, due to the nature of flame movement through the pack, the way that adjustments effect the flashing and fly ash pick up of the work inside, can instigate dramatic colours and surfaces.

Here I would like to discuss firstly, the variations in kiln design between the two kilns which I regularly fire, their similarities and differences, and how these construction elements affect the generation of micro atmospheres within the chamber. I will then speculatively analyse how these atmospheres can develop varying colour responses from clay bodies, ultimately culminating in a characteristic palette of surfaces ascribable to the different kilns.

The two kilns used regularly, are both anagama style kilns, consisting of a firebox at one end, a space in the centre for the packing of work and a chimney at the other end to provide the draw and pull the flames through the work. Both kilns have an 'organic', tear drop shaped chamber which tapers from front to back both downwards from the apex of the arch and inwards from the mid section towards the chimney base. Both kilns have side stoke ports situated in the side of the fireboxes in order to stoke long thin slats onto the coals. There are no side stokes further down the chambers, which means that all the heat required to fire the kilns needs to be produced from the firebox grates. I believe that this configuration allows the flame to have stronger continuity of passage through the work, and no additional disruptions to the flame paths are caused by the introduction of air further down the chamber. Having this configuration creates many different zones along the length of the chamber, which are constructed by placement of the work in the pack. It also creates the challenge of having to transfer heat from the front of the chamber to the back.

So, the elements present in the design of the kilns to control the chamber atmospheres are: 4 under-grate air channels and the lower fire mouth at the base of the door which when fully opened measures 23cm x 23cm (529cm2). This lower fire mouth is primarily used in the early stages of the firing to build the coals to a point where the fire can be sustained on the grate by stoking through the upper fire mouth. After this point the lower fire mouth is closed to a gap of 8cm x 10cm (80cm2). It tends to be left at this setting apart from through periods where we don't want the kiln to climb or want to encourage a sustained period of heavier reduction, at this point the lower fire mouth will be closed completely.

One of the major chamber differences between the 2 kilns is that the Wysing kiln has a stepped interior floor, where as the Loughborough kiln has a sunken firebox and a flat packing space. The Wysing kiln has a packing space of approx 110 cubic feet where as the Loughborough chamber is approx. 70 cu/ft. This floor pattern variation creates one of the biggest factors, which dictates the results that come from the kilns. The stepping up of the Wysing kiln floor combined with the descent of the back of the chamber arch creates a narrower exit towards the chimney. Although the entry flues to the chimney are the same dimensions in both kilns, the shape of the chamber concentrates the flames in the Wysing kiln far more. In addition the Wysing kiln has a flue, which extends for about a meter further up the hill before it then connects with a collection chamber at the foot of the chimney. The Loughborough kiln on the other hand has much more headroom at the back of the chamber due to the flat floor, and the chimney collection box forms the back wall of the chamber. Arguably it is these fundamental differences in the way that the draw from the chimney constricts and pulls flames through the back of the kiln that affects the entire gamut of atmospheres that appear to be generated in the entire kiln chambers.

Both kilns possess, passive dampers in the base of the chimneys. The passives are essentially removable bricks which when pulled out temper the draw of the chimney, slowing down the pull of the flame from the front of the kiln. In doing this, the path of the flame through the pack can be subtly altered. Areas of the pack, which may be receiving less direct flame and the alkalies carried in it, can be brought into more direct flame contact. It is also by this ability to change the speed of flame movement that the most can be made of the reaction between flame bourn alkalies and the clay elements. All these elements work together to construct the complex surface flashing and firing marks, which are so sought after in this type of firing.

The same fuel is used to fire both kilns. Scott's pine and Norway spruce slab off cuts combined with walnut slats. Pine generically has relatively high levels of both Calcium and potassium. The hard wood, augments the soft wood, (stoked through the upper fire mouth), by being introduced through the firebox side stoke holes. By introducing the hardwood in this way, it not only adds to the bulk of the coals by the fact that it burns slower with a shorter flame, but also by creating a criss-cross layering of embers, this allows the air to be pulled and pre heated more easily. Again, by layering the coals in this way more control is appointed to the 4 under grate mouse holes.

Weather conditions, which prevail through out the duration of a firing may also significantly influence the mechanics of firing. Both of these kilns are fired for three days, and over this period the atmospheric pressure can change quite radically. When high pressure is prevalent (usually characterised by dry bright weather), there is more oxygen at ground level and therefore more oxygen rich air will be drawn into the kiln through the air inlets. On the other hand if low pressure is in control (damp unsettled weather), there is less oxygen at ground level and therefore less oxygen rich air will be drawn in. When high-pressure systems are dominant the kiln will tend to be fired much more closed up with regards to the air inlets to compensate for the excess of oxygen. These factors may change through the duration of a firing requiring subtle alterations in the stoking patterns, air settings and passive settings to 'steer', the firing through these oxygen changes.

Finally the issue of firing duration may contribute to colour generation. However, having fired kilns from between 3 to 5 days and having now settled on 3-day firings, there was no discernable colour variations, which could be achieved over a longer period that I have not been able to produce over a 3 day firing. I would argue that it is through close monitoring of the atmospheres generated that has a greater impact, than the duration.




Having looked at the kiln design factors which can be used to influence colour generation and ash build up on the clay bodies, we should now look at how these elements can be used to produce the circumstances conducive to influencing the clays, other than variations in oxygen levels. It is widely acknowledged that many elements are carried through the kiln chamber and therefore the work, during a firing, and that these come into play within varying temperature ranges. Firstly and probably the most visible of these is the ash that is produced by the burning of wood. It may be argued that there are 2 periods within a firing that produce different qualities of ash. In the early stages the fire on the grate is small and the kiln chamber is still comparatively cool, therefore the draw from the chimney is less. This creates larger ash particles, which will tend to drift through the kiln in a gentle fashion being deposited on the horizontal areas of work such as the shoulders of forms and wider surface areas such as bowls and plates. Mixed with this slow ash is carbon in the form of the grey smoke that is produced in the early stages.

The build up of large particulate ash and carbon can be viewed through the fire mouths. The visible carbon is burned off the surface of work between 600 -700C, however the larger ash size appears to persist up until around 1100C. Once the temperature climbs above this at the front of the kiln. It has been observed that the pull from the chimney starts to increase and the flame path through the work becomes more horizontal. The ash that is given off by the fire appears to become finer due to more complete combustion and the more turbulent movement from combustion, which may break up the larger particles of ash, also, through changes in the chemical composition which will be illustrated later. From this temperature and above although there is to a degree, a continuation of ash, which travels up and is, then dropped further back in the kiln. There is also a significant movement of finer ash on to the surfaces of work orientated towards the fire. Below is a list of the elements present in a generic pine ash burnt at 600C by percentage weight:

Calcium 29.05
Potassium 16.24
Magnesium 7.03
Sulphur 1.07
Phosphorus 0.84
Manganese 4.04
Zinc 0.36
Iron 0.58
Alumina 0.47
Sodium 0.06
Boron 0.06
Copper 0.04

The elements that are also carried in the flame are fine alkalies of potassium and soda as well as small amounts of other minerals, as chlorides and sulphates. These alkaline materials are partly associated with the ash but are also present as free elements, which are disassociated from the courser ash at higher temperatures (Hence the visibly finer ash at higher temperatures). Anyone who has put their hand into a bucket of unwashed ash will attest to its alkalinity, evident by the burning sensation felt and also by the soapy feel of the water. The elements that are reactive with regards to flame carried materials as opposed to ash bourn, are those materials that become volatile at lower temperatures. In a study published in Biomass and Bio energy Vol. 4 No.2 in 1993 the authors studied the decomposition of wood ash (Including pine), at various temperatures. They postulate that ash formed at lower temperatures (600C) is high in calcium and potassium and that as the temperature is increased up to 1300C the mass of the ash decreased due mainly to the dissociation of potassium salts and the decomposition of both potassium and calcium carbonates, this dissociation also begins as low as 600C.

This would suggest that one of the most crucial periods in a firing is between roughly 600C and 1000C. Not only is this the temperature range at which the kiln would normally start to edge into its first cycle of reduction as a result of changing the stoking fully to the upper fire mouth, consequently promoting the fluxing and reaction of fine iron in the clay bodies. It is also clearly an important period for the production of flame born potassium that will also begin to react with the ceramic surfaces. It would also follow that pots which receive high levels of 'early firing' ash which has been produced at a lower temperature will acquire a fly ash patina which is higher in calcium (in ceramic terms used as a secondary flux), and that this calcium rich fly ash would at higher temperatures also begin to loose its combined potassium. It would therefore follow that fuel that is introduced into the fire box of an anagama at temperatures exceeding 1300C would form a much finer ash than that created earlier on in the firing due to this rapid disassociation of potassium and calcium elements from the ash bulk.

If this hypothesis is followed through, one would anticipate a visual and textural difference in the ash build up on pots at the front of the kiln that have received a larger level of early ash high in calcium and potassium, and those further down the chamber which may receive proportionally more of the ash produced at higher temperatures and therefore being more potassium rich. With calcium rich glaze one would anticipate that it would have a matter quality, and the more thorough the dissociation of primary fluxing potassium, then the more matt these surfaces would become. Pots further back in the chamber which have taken longer to heat up will have lost proportionally less potassium and also gained from the free potassium dissociated from the ash further forward, so arguably will possess a glassier fly ash deposit.

Other elements that become very important in the quality of fly ash glaze are the metal oxides, which are also delivered by the deposits of ash. The predominant metal being iron, although small amounts of manganese, zinc, copper and magnesium are also present. It is the reduced iron oxide which will give the fly ash its' characteristic green colour and when mixed with small amounts of phosphorus which is also present in ash, can give the ash glaze a 'chun' quality of light refracted blue. It must be remembered that at high temperatures iron in particular is a vigorous flux and will therefore aid in the movement of the fly ash glaze.

The surfaces, which have been discussed above in relation to the deposition of both calcium rich, and potassium rich ash are roughly the results that I have experienced on work being unpacked from the kilns. However the effects and colours are much more pronounced with the tendency towards darker colours with more variegated carbon trapping and alkaline responses from the Wysing kiln than from the Loughborough kiln, despite the firing cycles being very similar. The work coming from the Wysing kiln also tends to have a much matter glaze deposits on the pots at the front, with ash deposits becoming brighter towards the back. Where the ash build up has been thickest and run to the underside of side fired work, the drips have been contrastingly glassy.
The Loughborough kiln does not show such marked differences in the glassiness from front to back however there are subtle differences in the glaze. This may be a result of the Loughborough kiln fluctuating to a less heavy reduction due to its more open chamber towards the back of the kiln.

It can be deduced from the above that as well as kiln design having a significant effect on the colour generation from clays in this type of firing. Also, the time scale of climb in temperature and the wood burnt will have huge influence, as well as the final temperature both at the front and the back of the kiln. Due to the design of an anagama type kiln the back will always take significantly longer than the front to climb in temperature and therefore create quite diverse surfaces and complex actions and reactions in these different zones.

A generic graph of the relationship between the back and the front temperature climbs would show sharp peaks and troughs from the front, as the stokes causing the kiln to reduce as oxygen is sought to combust the fuel, depicted correspondingly as a trough, to a period of temperature climb as the fuel burns and gives off its energy creating a peak, then dropping to another trough which is slightly higher than the previous. This pattern would be seen to continue as the firing progressed. The graph for the back of the kiln would show a much steadier climb in temperature as the heat was gradually drawn further back into the kiln by the draw of the chimney. Therefore the work positioned in the middle and rear sections of the kiln are going to be exposed to very different time to heat ratios. This will also have an effect on deposited ash and the way that free potassium carried by flames from the front of the kiln are deposited and combined with the existing potassium and calcium etc. already deposited through the early stages of the firing.

It may be seen then, that the number of elements that come into play on the surfaces of work being fired in this way are complex. Not only do combinations of these elements produced from the firing itself influence the surfaces but the placing of the work itself is intrinsic to the way that these elements will interact with any given form. As has been described earlier, the surfaces of work that are exposed to direct flame impingement will pick up deposits of both forms of high temperature and low temperature ash as well as a continuous current of alkalies carried in the flame. On the lee of the work there is in a sense an even more complex set of parameters at play. It has often been remarked that the most dramatic flashed surface can materialise on these areas, and also on the undersides of side-fired work, sometimes also being influenced by the wadding used. It can be imagined that in these shielded areas the flame will cause greater turbulence as the draw from the chimney increases and the rate of flame passage through the pack intensifies. What effect this turbulence has, will also be affected by the pots, which lie directly alongside and behind. A pot may get direct flame from the front and have a quieter area behind, however, if the flame is also hitting the front of a pot slightly behind the first, then the flame will be deflected very gently onto the back of the foremost form. It is these subtle touches, which provide such seductive ephemeral surfaces.

Arguably one of the appeals of work fired in this manner is that it has 'faces'. It has a truly 3 dimensional quality and will have a different aesthetic appeal depending on from what angle it is viewed. If one looks into a viewing hole when the chamber is at good heat, after the stoke, the flames can be seen moving through the pack of pots and a deeper understanding of how fundamental the pack of the kiln is to a successful firing can be gleaned. Areas will be seen where the flame still moves slowly possibly due to work further forward shielding a section, in contrast, fast flowing areas will be visible where the flame rushes around and over, engulfing the work. The flame will ebb and flow like waves on a beach as successive stokes burn and recede. These multitudinous microclimates that are generated within the general firing scheme, are what create variety within the work.

Having looked at the different kiln configurations that are used, and discussed some of the chemistry behind the actualisation of clay surfaces, the question arises, how does one create these circumstances within the kiln chamber over a finite period of time? One would argue that there is no perfect time that a firing should last it is however the situations, which are developed and exploited that will produce the desired outcome. Many people fire for far longer than 3 days, but when firing this type of kiln the activity has to be fitted into daily life and it is hard for me to find periods of longer than 3 days in which firing can take place. So, therefore a set of parameters is set. Investigation then has to be carried out to tease the results out of this time frame, through subtle coercion of the kiln using the controls available. An intimate knowledge of the kilns and how they respond in any given situation can only be achieved over many firings. However, a theoretical knowledge of the aims can form the basis of an intuitive form of firing where attention is paid to all aspects of the kiln. It's sound, the rate of fuel being burnt, the degree of smoke and back pressure and the way that the flame is seen to move through the pack. All of these will give clues to what is needed next, and if an understanding of what is possible is achieved then there is no reason why the work coming out at the end shouldn't exceed expectations are even better, create new spheres of understanding and appreciation both of the forms and of the process that they have undergone. This information then forms the parameters for the next firing.

Firebox and grate description

Both kilns have roughly the same sized firebox (despite the Wysing kiln having a packing space that is approx. 40 cu/ft larger). Which consists of a grate with 4, 15 x 4 cm air gaps running beneath the grate. The under grate entrances are controlled by blocking with bricks or half bricks. The under grate entrances are also equally spaced along the front of the kiln allowing control of the air entering the kiln beneath the firebox coals, equally along the whole width of the firebox, this creates a maximum air intake of 240cm2. The grates are approx. 45cm deep by 120 cm wide with 4, 2cm air gaps coming up from below the grate and running the full width. A total intake capacity of 960cm2

Loughborough kiln dimensions

Grate size 18' x 45' 8102"

Length front wall to back flue wall: 93"

Width at front: 50"

Height at highest point: 35"

Width at back wall: 39"

Height at flue : 27"

Flue aperture: 1172"

New Wysing kiln dimensions

Grate size: 8642"

Length front wall to back flue wall: 112" (9'4")

Width at front: 48"

Height at highest point: 38"

Width at back wall: 32"

Height at flue: 15"

Flue aperture: 1442"

Wood ash composition as a function of furnace temperature. Misra K. Ragland K. Baker A. Biomass and Bioenergy Vol 4. No 2
Pp 103 - 116. 1993