home   info   gallery    articles    diary   links   contact    trivia

Water in Woodfired Kilns                                                        by Owen Rye

 First published in Ceramics Technical Number 10, May 2000, pages 17 to 22.

The unseen and often unconsidered factor in woodfiring is water. Many kilns built straight onto the ground absorb moisture that is released throughout firing. And the air contains moisture, less on a hot dry day and much more when rain is about. We all have ideas about the merits or otherwise of wet wood. What effect does this water have on our firings?

When the Different Stokes conference was being organised and I was asked for suggestions about topics for discussion groups, the first thing that flashed in lights was the subject of water in kilns. The reason was simple; I was very interested in the subject but not very knowledgeable and rather selfishly thought that someone else might have some experience that I could learn from. This came to be. Nisse Holstrom and Richard Bresnahan both had the experience I was seeking. Holstrom has been experimenting in Sweden and is building a new kiln, his Aquagama which has a sophisticated plumbing system installed underneath.

Richard Bresnahan’s talk at Different Stokes was for me the most interesting of the whole conference. Richard discussed research in Japan starting with archaeologists excavating a C12-13 Tokoname kiln in Japan which appeared to have a water supply built in underneath the floor in the form of large water storage jars that could be filled with water. Later the Idemitsu Museum sponsored research into a replica Kamakura kiln which was fired in an attempt to replicate the original medieval pottery. Over a series of firings this was achieved, suggesting that a similar firing cycle to the original Kamakura and later Echizen ware had been developed. Subsequently experience with a kiln on Tanegashima island further demonstrated that a distinctive colour palette and surfaces could be produced through water introduction at the end of firing. Richard went on to build a massive kiln in Minnesota which incorporates various methods of introducing water during firing, and his knowledge and experience in this area is extensive.

My interest in the effect of introducing water during firing started, as do many obsessions, by accident. Around 1991 Chester Nealie had a contract to provide an installation for a world expo in Spain, but had no kiln. So I suggested he fire in the Gippsland anagama. The work was extremely important to Chester; the firing had to be a success. I decided to wash the ash and dust from the previous firing out of the kiln with a hose, so the floor was clean but wet. Then it seemed like a good idea with our perverse woodfirers logic to take a big gamble and saturate the whole kiln with water before the firing to see what might happen. What did happen was a superb firing, with rich red flashing colours under wads and rich colour throughout.

I have been experimenting for eight or nine years since then with the effects of introducing water into an anagama kiln. The aim has been to:
enhance colour of glazes via extra reduction
enhance flashing colours and colour under wads
matt glazes via increased reduction especially during cooling

Water can be introduced by various methods. My tendency is to use the simplest means, with no equipment to go wrong during firing. The simplest way to introduce water into a kiln is to fire a kiln built on wet ground. The water is released throughout the firing progressively as the heat penetrates. I use this principle in the Gippsland anagama in which the floor is a single layer of bricks placed directly on the ground, by simply soaking the floor of the kiln with a hose for one or two days before stacking. This conveniently allows me to wash all residual ash and other waste out of the kiln at the same time so I can stack the kiln without the need for a dustmask. I am abnormally opposed to breathing in ash dust so the wet kiln keeps me happy in my work.

The slow release of steam continues through the firing as heat gradually penetrates further into the wet floor of the kiln and the ground underneath. This steam seems to have the desired effect on glazes and on flashing colours. It unfortunately has some other very definite effects during firing. One is to move the firing towards very strong reduction; in fact so strongly in the early part of the firing (below 1000C/ 1832F) that I have had the kiln completely stalled by carbon whiskers building up and blocking off all flame passages. Consequently the setting for a wet kiln has to be more open than for a dry kiln, more like the kind of setting I would use in a dry kiln if I wanted to do an oxidised firing. The stoking pattern also is more like that which would be used for an oxidised firing, with longer intervals between stokes allowing the wood to burn down. And even though the kiln is fired in heavy reduction (readings up to 0.8 on the oxyprobe) it is possible to achieve higher temperatures more easily than it would be firing a dry kiln.

The oxyprobe is an essential piece of equipment for these experiments, because the familiar signs of reduction learned over many years are not evident when water/steam is present. For example, when at higher temperatures the oxyprobe is indicating a high level of reduction (say 0.8) the interior appearance of the kiln suggests to me a neutral or slightly oxidising atmosphere, as it would be if steam was not present. One particularly striking incident occurred during 1999 at the end of an anagama firing. My practice in ‘normal’ (that is, waterless) firings has been to end by stuffing the firebox with dense wood and almost completely sealing the kiln. The result is a vast billow of smoke accompanied by a fierce reduction flame from the chimney top. On this occasion, for no good reason other than that it was there, I decided to use almost green blocks of pine which were very wet from sitting in heavy rain for two weeks. After stoking we went outside to watch the show – and nothing happened. No smoke. Clear chimney. Accompanied by that feeling that something bad was about to happen. Nothing did; and some very satisfactory work emerged from the kiln a few days later.

In this same firing I remember feeling dissatisfied with my regular firing companion Graham King. I thought he was getting lazy and leaving far too much time between stokes, until he showed me that the oxyprobe readings were indicating serious reduction when the kiln appeared oxidising. Graham was much more alert than I was at that point, and his firing method was quite appropriate.

I have spoken to Max Murray, the inventor and maker of the oxyprobe. Murray told me that the oxyprobe measures the oxygen differential between the kiln atmosphere and the outside air, and so it should not matter what agent is being used for reduction (carbon monoxide or hydrogen). That is, the probe will give an accurate indication of the level of reduction regardless of the agent creating the reduction. Max also suggested, in relation to water introduction, that one way of preventing the cooling effect of water on the kiln might be to introduce an alcohol/water mix (about 15% methanol/85% water for starters), so the alcohol as an additional fuel will maintain temperature. I may have to rethink my rule about no alcohol near the kiln during firing!

I am becoming more dissatisfied with the wet kiln/wet floor before stacking method of water introduction because of the problems in the early stage of the firing: excessive and particularly nasty smoke, and excess carbon stalling the kiln. The results after firing are worthwhile, so I am looking for other methods to introduce water. I am building a new anagama at Boolarra South, so it seems like a good time to experiment. I liked Holstrom’s idea of plumbing underneath the floor of the kiln, so I have incorporated galvanised steel pipes under the brick floor. The pipes are drilled with many 1cm holes so that when they are filled by pouring water into their upturned ends the water will gradually drain under the floor bricks and penetrate into the kiln. About 2-3 cms (one inch) of loose gravel under the floor bricks should allow the water to spread somewhat. I am hoping that if or when the steel pipes rust away that the loose gravel will still allow water to flow under the floor. I’m also hoping the whole thing does not explode from steam pressure. You will read about it if it does, I guess.

This method, if it works, should allow me to fire the kiln “dry” until a temperature of 8-900C ( 1472-1652F) is reached, and then introduce water at whatever temperature and rate seems to work best.

Another obvious way to introduce water during firing is to use wet wood (not unseasoned wood, which is something different) at various stages of a firing. The temperature will drop gradually due to cooling of the ember bed and slow combustion of the wood but reduction will be intensified. In the anagama I generally only find this to be a useful technique if the wet wood is used to produce reduced cooling results. I was intrigued by Richard Bresnahan’s method for using wet wood in his large kiln. Spaces between stacks in the kiln allow him to stoke long large logs so they stand vertically and burn from floor to arch. This technique would tend to distribute the water vapour and its products throughout the entire kiln rather than in localised areas.

Obviously there are various other potential methods of introducing water, such as spraying or drip feeding it into the firebox, or even simply throwing it in by the cupful. I might try dripping it through an alumina tube directly onto the pots to see what happens. As more potters experiment with these techniques we will gradually develop new aesthetic approaches incorporating the results.

The devil, as they say, is in the details. The temperature and time ranges for introducing water during heating, and especially during the cooling phase, will be the subject of many experiments. My experiments have not been very carefully controlled, more just a ‘try it and see what happens’ approach. An anagama only fired a few times a year is not a good tool for systematic investigation because it takes too many years to achieve full understanding of the variables. Some of my postgraduate students firing gas kilns for quick results, are looking at water introduction in a systematic way – as well as the effects of reduction at various stages of heating and cooling. Ray Cavill’s experiments on reduction have shown some extremely interesting results (see Ceramics Technical No 9, 1999). Susan Scott is experimenting with introducing water as a mist but results are not yet available. Gail Nichols has completed her experiments with introducing water into a soda- glaze kiln, and her results are summarised elsewhere in this issue. They are somewhat idiosyncratic in that she has been interested mainly in how water introduction effects the specific soda glaze she has been developing. They are of general interest in that she has found that water introduction at various phases of firing has a very definite effect on the final glaze quality.

One of the findings by both Gail and Susan, carefully measured and recorded, which everyone should consider is that the weather does matter. Their attempts to obtain results under carefully repeated conditions were sent astray by rain. A wet day does very definitely give different firing results than a dry low humidity day. The thinking of the old traditional potters on this subject is well founded. Wet days, more reduction, slower firing. So the simplest way to play with water in your kiln is to wait for rainy day to fire.

Thinking about traditional potters reminds me that there is a very considerable technical literature waiting to be trawled for directions for experiments; not necessarily directly relevant to woodfiring but suggestive nonetheless. The literature on fuels and combustion contains many references to the chemistry of water use to enhance fuel values, or to increase reduction, and a look through 19th and early-mid 20th century literature may be very profitable. As an example water gas was produced by blowing steam through a coke or coal bed previously superheated by blowing air through it. The water reacted with hot carbon to produce carbon monoxide and hydrogen, subsequently used as fuels. To me this suggests that carbon (as embers, or carbon ‘spiderwebs’ in the anagama) is an essential factor in the water equation.

It seems from Richard Bresnahan’s comments and from some of Nisse Holstrom’s reading (summarised in the Different Strokes conference proceedings) that water introduction into woodfired kilns was probably common in the past in Asia. Some of the more scholarly among us may find a worthy research topic looking for literary or pictorial evidence of this. If there is someone out there interested in doing an MA to examine the historical and technical practice and consequences of using water in kilns, please contact me. There is no doubt that water introduction has aesthetic consequences, and I am hopeful that as more potters experiment we will begin to understand and use the possibilities in new ways. It is quite clear to me now that colour development and glaze matting are two aesthetic qualities strongly influenced by water in the kiln.