A+D Article-Ray 1991

This rather unusual portfolio addresses the issues of energy, labour, cost and ecological viability of our building construction techniques. In particular, it sets one thinking about the energy that goes into the making of our buildings - energy that is substantially higher in the Indian context than the energy consumed in the use of our buildings. The designs can, of course, be e:njoyed purely as excellent sculpture, but doing so to the exclusion of other view points would be entirely missing the significance of the work.

Ray Meeker and his colleagues have been pioneering in situ brick firing technique & for a number of years in Auroville. On the one hand this work can be seen as one possible way to stabilize a raw earth structure, a problem that really has not been solved satisfactorily. On the other hand it can be seen as an attempt to reduce the firing and transportation energy consumed in conventional fired brick construction. But on the whole it is probably best to see these experiments as representing a branch of building technologies sufficiently distinct from both raw earth or improved earth construction techniques on the one hand and from fired brick building on the other so as to merit exploration in its own right.

In modem times, reports of in situ fired structures have been trickling in from the Middle East where they use oil-fuelled blow torches to stabilize raw earth buildings, a shocking prospect in our context. Meeker's work is much better suited to our environmental con- text, utilizing biomass fuels which are, at least in principle, renewable. In practice, of course, this may still entail a heavy price in the short and medium term, unless married with sound afforestation practices.

At first sight it seems that the energy savings in the techniques described are the result of firing only the finished masonry face. But this is not so. As Meeker candidly admits, a kiln designed for efficiency is not the greatest of places to live in, and a compromise has to be reached between efficiency of fuel burning and spatial requirements and use. Also, since the firing has to be from inside but the heat must reach outside in order to stabilize the face that matters most, the entire mass of the building has to be fired. All this results in a structure whose energy cost is actually somewhat higher per unit volume than that built in efficiently fired kiln bricks, except that there is a substantial saving of cement by its virtual elimination in the building, especially in the mortars.

Therefore, the justification for the technique really centres around the other fired products that can come out of the 'house-kiln' .These pottery products can be used in the house itself, or sold. If the technique were to be propagated on a larger scale, it is in this area that the implementors would have to concentrate, and this probably makes the idea fairly untenable for sale to the bureaucracy .Unfortunately, integrated concepts addressing a range of problems simultaneously, such as employment, local self-reliance, energy, distribution of natural resources, and housing, as this one does, tend to slip between the cracks in diverse institutions, and be forgotten.

There is one main problem that this technique does not solve, and that is the question of large scale availability of suitable soil and the agricultural consequences of using this soil for building, though the impact in this area is not likely to be any more severe than that of fired brick and most forms of raw earth construction.

Whatever the controversies, reservations, and arguments for and against the technique of in situ firing, (and there are many), it is undeniable that the work described in these pages is of inestimable value in terms of both its scientific and its artistic merit.
Sanjay Prakash

A good deal has been said over the past fifty years on the need for a revival and/or upgrading of the mud building technique in order to produce a durable 'low cost' house. The call, strident in the past decade as housing demand increases, energy costs rise, and natural resource reserves seem to be on the decline, has yet to yield a totally satisfactory solution. To be sure, there has also been strong opposition to mud as a viable building material in the twentieth century. The resistance, not unjustified, stems from technical, cultural, and no doubt, political and economic prejudices. Without getting bogged down in the mud versus cement rhetoric, it should be sufficient to say that any technique which could produce a reasonably low cost solution to today's housing shortage is worth serious investigation.

The Iranian architect, Nadir Khalili, has pioneered a method for 'stabilizing' mud structures. In his book Racing Alone (Harper and Row, 1983), he relates his five year quest for a technique to improve the village house which culminated in the firing of an existing house and, subsequently, the building and firing of a ten-room school. To me as a potter with a university background in architecture, living in India -a country with
both an acute housing shortage and a tradition in mud building, Khalili's experiments hold a special appeal.

In 1985, with the assistance of Ian de Rooden, a Dutch ceramist funded by the Dutch government to come to India as a consultant to the project, I began a series of experiments which hopefully, will lead to both a technically sound and economically viable stabilized mud structure.

Essential Process
Fundamentally, the process for building and stabilizing a mud brick structure by fire is as follows:
- Build a room/kiln in unbaked mud brick. The roof can be either a dome or vault.
- Fill the room/kiln with a product; bricks, tiles, drain pipes, etc.
- Fire the whole mass as a kiln to between 900°C and l000°C (or to the proper temperature for the locally available brick clay).
- Remove the product and use a portion for finishing the kiln as a house; tile the roof and floor, build a small compound wall with bricks, install window jalis, toilet pans, chullas, anything appropriate to a home that can be made in fired clay.
- Sell the remainder of. the product to recover as much of the building cost as possible.

If product sales cover their own production cost, including fuel, then no fuel cost is accountable to the structure. In other words, the cost of the fire-stabilized structure is identical to the cost of the mud brick structure.

Basic Considerations

Six test structures were built and fired in the first three years.
The first six tests were conceived for basic experience in building mud brick vaults without centering (the Nubian method) and becoming familiar with the behaviour of large mud brick structures during firing. Further, it was necessary to determine whether a 15cm thick vault could be thoroughly fire-stabilized, with how much fuel, and what kind of post-firing treatment would be effective.

Structure 1 A single vaulted room (2m span, 1.35m rise 3m long ~ on 1.65m high walls), was filled with 4,500 bricks and 1000 small 1 tiles which were laid over the vault after firing. It was fired with ~ wood to 960°C in a 40-hour cycle. The exterior of the vault was ~ insulated with a mixture of clay, ash and sand in a layer 10cm thick.

This first test yielded product bricks of remarkable fired quality, far superior to the locally available brick. The vault, however, was really not sufficiently stable on the exterior. The walls, though well-burnt on the inside, were absolutely un- touched on the exterior .

The most significant reaction of the structure to the firing process was the separation of the gable wall from the vault. Interestingly, this in no visible way affected the integrity of the vault. This separation is caused by differential thermal expansion. As the structure reaches a temperature of 850° to 1000°C, there is significant expansion of the walls, but only on the inside. The exterior, being cold, does not expand. This causes the walls to bow and lean out of plumb, separating altogether from the vault. Though the gable walls move back on cooling they do not return to their original plumb position.
Structure 2 A single valuated room 3m span, 2m rise, 5m long, on 2m high walls. Insulation of the vault was as in Structure 1, and the aim was to test a structure with both a significant carpet area and height.

As before, the exterior walls remained untouched by fire. Though a number of methods for protecting a mud wall with plaster exist, I opted for a composite wall in this test, which being very thick (45cm), was built with a fired brick on the exterior. Though this does raise costs somewhat, I believe the expense will be justified by the increased life span of the building.

This structure held 18,000 bricks and was fired to 940°C in 48 hours. Predictably, the gable walls bowed outward. The exterior of the vault, though somewhat improved in terms of fired quality, still sho':-led large areas of unstable brick.
Structure 3 A single valuated room, 2.25m span, 1.5m rise, 5.5m long on 2.14m high walls, insulated with two courses of fire brick.

This time a significant change in construction technique was examined. In order to solve the problem of plumb, the gable walls were eliminated altogether and added in fired brick after the burning of the building was completed. There are advantages as well as disadvantages to this. The end walls and their foundations can be much less massive, and built in completely stabilized brick. However, this does necessitate the use of some kind of centering as there is no end wall on which to lean the vault as is essential to the Nubian technique. Vault 3 was constructed over a moveable section of centering lm in length. Though one can argue about the extra expense and complication of using centering, in fact, it is an excellent way to maintain the catenary curve necessary for the stability of the structure.

For firing, the open ends of the structure are temporarily filled with unfired brick. Structure 3 held 13,000 bricks and was fired for 78 hours. Though fuel consumption increased slightly, relative to the first two tests, the long, slow-firing cycle penetrated deeply into the roof. The vault exterior showed much less unstable area.

Structure 4 A single vaulted room, 2.25m span, 1.5m rise, 5.25m long, on 2.14 high walls. This time the insulation used was a combustible mixture of cow dung, straw and clay. Four technical alternatives were attempted and all proved valuable.

The side walls in the first three buildings showed a movement similar to, though less pronounced than, the movement of the gable walls. The movement was monitored in Structure 3 and in Structure 4 the walls were built out of plumb-leaning slightly inwards. After firing, the walls had moved, yet remained just out of plumb -inward.

The vault was built with a combination of techniques: A 75cm section was built over the frame approximately midway between the ends of the side walls. The vault was continued with the Nubian method, leaning bricks against the 75cm section. The frame was again used to finish the vault at the ends.

The Objective in the first three structures was simply to insulate the vault. The fourth vault was covered with a combustible mixture of cow dung, straw, and clay which served as insulation up to a point, then ignited to contribute to the stabilization of the exterior of the vault. This combustible insulation continued to burn slowly for two days after stoking of the fire was completed. As a result, vault 4 was completely
stable, inside and out. I would like to point out that the same result could be achieved by firing the structure to a higher temperature on the inside, but as our local brick clays will not tolerate much over 950°C, this is not possible without weakening both structure and product.

The fourth technical improvement realized in this attempt was in terms of the product. The product brick for the first three tests was made by the slop-mold technique, on the ground. As a long, slow-firing procedure and, by local standards, a higher ul- timate temperature are necessary for properly firing the structure, much more fuel is used than in the traditional skoves of the South. The fired product brick is much superior to the local product; but it looks more or less the same and can be sold at only a slightly higher price. The product for this test was well- formed table-molded brick. Though these cost more to make, we did get abetter return on the overall cost of producing the building as all other process costs remained unchanged. While the sale of these bricks in no way covers building costs, it does indicate clearly the potential of product development.

Structure 5 A single vaulted room of 3m span, 5m length, and a total height (vault + wall) of~.14m. Containing a product load of 12,000 table-molded bricks, it was fired to 945°C in 84 hours, using 10 tons of dry casuarina as fuel. The vault was insulated with a combustible mixture of rice husk, cow dung, and clay as were all subsequently fired structures.
Structure 6 Again a single valuated room, 3m span, 6m long and 4.65m high. This room contained 19,000 product bricks and was fired in 77 hours to 920°C. The firing consumed 7.37 tons of casuarina and 1,000 bundles of casuarina malaar (the top and large branches of the tree, including the needle-like leaves). A bundle of malaar, depending on size, has roughly the calorific value of 7.5 kg of casuarina wood (4950Kcal/kg).
In this high vault, a concrete slab was cast after firing, giving a partial second floor and increasing the carpet area significantly for little additional cost. This demonstrates the potential for incorporating conventional building techniques in a fire-stabilized structure.

Architecturally, tests 3,4,5 and 6 have been combined to form a single house. Connecting the four vaulted rooms by a series of unplastered brick walls (using bricks fired in the structure) creates a rich play, spatial and textural, of open courts, corridors, and covered areas.

Product Development

Product development in the broadest sense has two main objectives: The first and most immediate is to generate income from the process and produce as many building components of some kind as possible. Second, though both ambitious and extremely optimistic in outlook, it must at least be postulated that a small scale, labour intensive, rural building industr, producing a wide variety of fired clay building components, is a possibility. The village potter, whose livelihood is now threatened by mass produced plastic and aluminium vessels, and whose knowledge and skill with earth and fire are no longer fully utilized, holds a great potential for creating a rural clay- based building industry.

Given the number of potential products -table-molded bricks, glazed facing bricks, terracotta and glazed tiles, drain pipes, jalis, toilet pans in glazed earthenware, terracotta bio-gas plant and burner, decorative and ritual elements of the house such as altar and depam niches, smokeless stove, terracotta refrigerator, water filter, filter candle -it would seem that practically the whole dwelling system could be done in fired" clay.

It can be argued that it will not be possible to compete with existing industry in terms of either product quality or cost -that this is a dream. This may be true, though if a number of products and processes are developed that are appropriate to different economic conditions, these objections might be over- come. For instance, a project in a backward area, far from industrialized urban centres, will not have to compete with mass produced clay building products as they generally do not reach these areas.
Producing with labour intensive methods, the products will, no doubt, be of somewhat inferior quality, but if quality is sufficient for utility, then a local market -albeit a limited one- should exist. In areas adjacent to urban population with broad market potential, if local raw materials are good, a more sophisticated process and product could be considered.

A House in Auroville

An experimental house, built and fired in Auroville in 1988, was a huge leap towards firing complex systems. This structure, a dome surrounded by four vaults, had a floor area of 65sq m in the firing zone. After firing, a fifth vault was added (2.5m long and 3.5m high) as an el\try, in fired brick and lime/sand mortar.

A partial RC loft was cast in the highest vault. The building is set 75cm below grade, which minimizes the need for buttressing the vault and maintains a low profile on a site which is rather prominent. The total living space is 74.32 sq m and has been occupied for the past year. Even without fans, it is notably cool.

The structure was stacked with 60,000 table-molded bricks, 2,000 tiles, (which were used to floor the house after firing), a number of decorative drain spouts, and several terracotta toilet pans.

It was fired for 4 days with 23.5 tons of casuarina wood and 2,400 bundles of malaar. The four vaults were fired simultaneously, cross-draft towards the dome, which was then finished as a down-draft system. A down-draft chimney placed in the center of the dome rose 2m above the exterior height of the dome. The temperature reached 900°C to 980°C at the five thermocouple points, but doubtless varied much more through- out the entire system -probably 800°C in the cold spots to 1100°C where bricks began to melt.

Towards a Low-Cost Solution

The ultimate goal of this series of experiments has been to develop a technique which, in the right circumstances, can be one solution to building a low-cost, high quality , climatically appropriate house.

The next building, a 'model' village house, was a single vaulted space 3m span, 7m long and 3.Sm high. It was divided into three areas after firing: (a) a walled entry court, partially covered by the vault (1.S6m x 3m), (b) a living/sleeping area (3.88m x 3m), and (c) a kitchen area (1.S6m'x 3m). A walled courtyard in the rear contains a WC and a bathroom (each O.92m x 1.3Sm). The remainder of the court is open to sky (1.9Sm x 1.64m). All walls are plastered inside and out with lime/sand mortar. The vault is waterproofed with Scm of brick jelly and lime covered by thin fired brick tiles. Cement mortar is used at sensitive joints -on cornices where vault meets wall. The five doors are in ferrocement. Light and ventilation are provided by narrow vertical openings above and to the sides of the doors. The floor is terracotta tile fired in the building. Electrical connections with four light points and two plug points are provided.

The house was constructed by a local contractor. I supervised the building of the mud vault and produced and fired the product (and incidentally the building) independently of the construction cost of the house, which ultimately came to Rs 861 per sq m of built-up area.

A Site-Specific Process

A typical single-vaulted structure will require 6,000 to 7,000 bricks to build. The same structure will contain 12,000 bricks as product. It is, of course, normal to move a considerable amount of material for building a house -but 12 to 14 truckloads of earth for the superstructure of a 14sq m room? Purchasing raw bricks from the local brickmaker at Rs 200 to Rs 250 per 1,000 is of little help. Good brick clay on or very near the site is essential to realizing the full economic potential of the process. Mud is generally considered to be an abundant resource (it definitely is when it rains on a unfired clay structure), but clay suitable for making fired bricks is not to be found everywhere.

BURNING EFFICIENCY COMPARISONS
TEST NO. DATE

STRUCTURE

 L W H

KILN TYPE PROUDCT FIRING TIME TEMPERATURE FUEL CONSUMPTION MJ/1000 PRODUCT BRICK MJ/1000 BRICK INCLUDING STRUCTURE1 VALUE OF PRODUCT COMMENTS

5

1987

VAULTED ROOM

5m x 3m x 4.14m

Up-draft

Table-moulded

brick-12,000

84 hours

945°C

Casuarina 10 tons 17,270 12,300 Rs.10,000 Considering product brick alone, 5 and 6 are roughly equivalent to standard up-draft kins
 

VAULTED ROOM

6m x 3m x 4.65m

Up-draft Table moulded brick-19,000

771/2 hours

920°C

Casuarina 10 tons Malaar 1,000 bundles 16,292 11,680 Rs.15,800 Efficiency is gained from cross-draft in vaults and a well pre-heated dome. Dome temp = 600°C on flue gases from vaults
 

VAULT 1

4 x 3m x 3.5m

VAULT 2

4.5m x 3m x 3.5m

VAULT 3

5m x 3m x 4.5m

VAULT 4

6m x 3m x 4.5m

DOME

5m dia x 5m H

Fired simultaneously cross-draft into dome

Down-draft

Table, moulded

brick-60,000

Floor tiles –

2,000

Toilet pans – 4

Decorative

drain spouts – 35

108 hours

900-980°C

Casuarina 23,4 tons

Malaar

2,400 bundles

13,270 9,100 Rs.75,000 Small up-draft kiln is very inefficient, but product diversity yields more value
 

VAULTED ROOM

6m x 3m x 3.5m

Up-draft

Table-moulded

brick-8,000

Floor tiles – 1,000

Hollow roofing elements – 1,000

Hollow wall elements – 60

Terracotta

mural – 18ft2

48 hours

940°C

Casuarina 6.6 tons

Malaar

600 bundles

19,732 14,250 Rs.16,000 Small up-draft kiln is very inefficient, but product diversity yields  more value
 

VAULT

4m x 3m x 2m

Cross Draft

Table moulded brick – 3,500

36 hours

930°C

Casuarina

2 tons

11,850 8,457 Rs.4,530 This was the test for an efficient village kiln
 

3 VAULTS

6m x 3m x 3.5m

Semi-continuous cross-draft Table moulded brick – 35,000

96 hours

940° C

Casuarina

10 tons

Malaar

1,400 bundles

12,193 7,900 Rs.43,500 This system can be vastly improved by a change in flue positions.
  1.  These figures are somewhat misleading as structure bricks are well fired on the inner wall face and are progressively less well fired through the wall. One brick depth is considered to be in the firing zone in the walls. However, the figures for the cross-draft 3 vault system are real, as walls are fired from both sides.
Fuel Consumption in Standard Kilns    
TYPE OF KILN MJ/1000 PRODUCT BRICK COMMENTS
Spove2 16,000  
Up-Draft 16,000  
Down-draft 15,500  
Hoffman 4,500  
Habla 3,000  
CBRI Biomass3 5,800  
Local Skove 5-6,000 Fired quality so poor as to make comparison meaningless.

2 Figures for the first five kilns are taken from Small-scale Brickmaking, Technology Series, Technical Memorandum No.6, International Labour Organization 1984.

3 Satya Prakash and FU Ahmed. “Use of Rice Husk as Fuel for Building Bricks,” CBRI, Roorkee.

 

The tenth test, the first building to be done in nearly ideal' conditions, is a series of three vaults on common walls. It was fired as a semi-continuous chamber kiln.
The on-site clay is of sufficient quality (just) for brick making. 0\11 structural and product bricks were made at the site.

Fuel Economy: The Ecological Crunch

The facing table gives a comparison of fuel consumption for six fired structures and for seven conventional firing systems. These figures are reasonably accurate, and though peak temperature md soaking time vary somewhat, the trend is quite evident and IS consistent with the general knowledge on firing systems.
The up-draft house and up-draft kiln data are nearly Identical and. it should be noted that up-draft kilns are notoriously inefficient. Compared to the best firing systems available -the Hoffman, Bull's Trench, or Habla - there is certainly no future for firing houses up-draft (or down-draft). However, the :ests do show a remarkable progress towards reaching an acceptable level of fuel consumption: From 13,400 MJ to 7,900 MJ in the 3-vault system is not an unimpressive advance.
There is certainly a limit as to how closely continuous kiln economy can be approached. There is a trade-off because this structure is not only a kiln -after firing it becomes a living space. I have heard of chamber kilns being used for shelter in Europe during and just after World War II, but I do not think the most efficient kiln will make an ideal living space.

In the 3-chamber test, the vaults are higher than they are wide, which slows the draft. Heat tends to stagnate at the top of the vaults. Interestingly, heat loss to the structure, an unfortunate waste in ordinary kiln operation, is an advantage when firing a house, more especially when the loss is through the roof.


A Problem of Scale

Long, multichambered structures are unquestionably most efficient. It is conceivable that a 10-vault system will prove economical. But even ten vaults, by no means a large project, will require producing some two lakh bricks, inclusive of structure and product. This is way beyond the means of individual families and probably necessitates government subsidy and support.

It must also be remembered that until the building is fired, it is all mud. Large-scale mud building projects are extremely vulnerable to 'unseasonal' rain,more especially when a large quantity of mud product has to be protected as well. Process infrastructure would include a significant and not inexpensive area of canvas tarpaulin. This need not, however, invalidate the process. It probably precludes it from use on very large projects. In fact, many villages in India are small, and I do not believe that each and every building technique evolved must necessarily be appropriate for covering hectares of land with literally thou- sands of houses. Traditionally, in rural areas, building and planting respects the cyclical monsoon pattern. It is well-known that large-scale government housing projects contribute a very small percentage to the total housing stock, and this trend will most likely remain unchanged in the foreseeable future. What of the possibility of a single family building and firing its own mud house?

On the Auroville project, the masons were quickly to appreciate that they could make their own mud bricks and construct the mud building for very little more than the cost of their own labour. But it was equally obvious to them that they would never have the capital necessary for firing. I suggested that they could build the house in mud and I would make the product, fire their house and recover my cost from product sales. This, though feasible enough, does not solve the problem. From the chart it is obvious that a small room fired up-draft is prohibitively inefficient.

In a test to develop an efficient, inexpensive, fuel-burning kiln for the village potter, I have constructed a simple catenary vault on a low plinth of fired brick. The kiln, built in unfired brick, is 4m long and has a chimney at one end in fired brick. It is fired cross-draft to 950° C and requires 11,800 MI per 1000 bricks.1 For such a small kiln, it is very economical.
This is not unlike a small version of the CBRI bio-mass kiln, which uses 5,800 MI per 1,000 bricks. In a larger catenary vault system, fired longitudinally, it is conceivable that a figure of 6,000 MI can be reached. The next test, scheduled for Auroville in 1990, will demonstrate the longitudinal draft system in two connected vaults. If Habla (or zig-zag) efficiency can be approached in this test, then fire stabilizing a single small house will indeed be economically viable.
The village potter could easily build a kiln with this method and, after understanding its principle through a number of firings, put a similar vault on 1.5 m walls and fire a room to live in. He would then become the equivalent of the local mason as a maker of houses. He would use the house as his kiln for firing floor tiles, jalis, chulas, pipes -any number of items which could contribute to the stock of locally available building materials. Transfer of technology, always a stumbling block, might thus be handled with a minimum of difficulty.

Factors in Cost

If on-site or very near-site clay, suitable for making fired bricks is available, material cost for the superstructure will be negligible. Foundation costs are somewhat high, as walls are thick (common walls 45cm, end walls 35cm with buttresses) and the vaults fairly heavy .

As the current test is not yet complete, I cannot give an adequate accounting for the three-vault system. With on-site clay being used for all brick, structure and product, the original estimate looks quite good at Rs 376 per sq m of plinth area. With the inevitable overruns, it should remain within Rs 484 per sq m, which is very reasonable, considering the quality of the building.2 (The table on p88 gives the final results)

This process is labour intensive. Most of the work is unskilled -moving bricks, loading/unloading the structure, adding cow dung/rice husk/clay over the vaults, removing it after firing. The firing labour itself, continuous for 60 hours or more, even on small houses, is exhausting. This can be seen as positive -many people get work, and a high percentage of cost goes to labour rather than material, or negative -labour cost is high even though wages are low, as is productivity .Evidently, labour intensive does not mean work intensive.

All fuel and firing cost can be considered accountable to the product, or a percentage can be added to the structure (house mass relative to product mass).

Post-firing finishing costs will vary depending on treatment. All structures will need proper waterproofing over the vaults. This is most important. The structure is stable, will not slake in the rain, but, like a fired brick, it is porous. The vault will absorb a large amount of water, increasing its weight tremendously, and of course, will eventually leak. Exterior walls should be plastered with cement or lime-base mixtures.

Interior walls can be treated according to budget: cement or lime plaster, mud plaster, or whitewashed with no plaster at all. Floors can be tiled with fired product. Windows can be filled with fired jali elements, and drain pipes, toilet pans, smokeless chulas installed.

Limits
Before concluding I must add a note on vaults built in the leaning, or 'Nubian ' technique. This structure is susceptible to cracks either due to unresolved tension at the top of the curve or to thermal expansion in hot climates. The latter generally appear along the weakest vertical joints. Many of the vaults.I have tested have developed post-firing cracks. I have been concentrating on developing the firing economy of the system, and as yet have not had time to worry about these cracks. Mud vaults of this type, cracked and otherwise, have been standing for centuries.

Architecturally, there are severe limits inherent in this process: Roofs must be vaulted or domed, and interior spaces must not be complex in order to ensure adequate draft during firing. The Auroville house, though not architecturally complex, was certainly not a typical kiln and posed real questions as to whether or not it could be fired in one single operation.

Far more serious will be the problem of maintaining ade- quate construction standards. Criteria can be established,even codified for proper building procedures. Fire (or fuel) is the cement of this process. Either the factor of the contractor cutting costs, or the poor villager unable to afford sufficient fuel to fully stabilize the structure, could lead to a structural failure with serious consequences.

Incidentally, for a large project, a very big hole in the ground must be incorporated into the design concept. A public water tank may be desirable, or perhaps a commercial fish farm should be considered, providing a good local protein source. Comprehensive planning must in any case be done for mass housing projects. Let us have no more rows of 'low cost' houses deteriorating in an environment more suggestive of an unwalled prison camp than a village community.

In Conclusion

Fire-stabilized mud building, when perfected, will no doubt be limited to a very specific set of circumstances. However, where these conditions are met (good on-site clay and a local fuel source), it is reasonable to conclude that a house of relatively high quality can be built for a very low cost. Being extremely labour intensive, the fired building process will create jobs and, using almost exclusively local materials both for house and fired product, will help to decentralize the construction industry. Hopefully, the next experiment will bring this process into a new phase: transfer of technology .

The Eleventh Test 1990

The Auroville Information and Reception Centre, a project of the Centre for Scientific Research, Auroville, was conceived architecturally as a demonstration of building with stabilized soil blocks and other appropriate technologies, namely solar and wind power generation. As a part of the project, I was asked to build two mud houses (room, kitchen, toilet, bath, veranda) and to stabilize them by fire: one house to be used by the watchman; and the other for a small exhibit on how the houses were made. A total of Rs 20,000 was allotted for the two houses. For me, it was an opportunity to demonstrate the potential for income generation in this very small sys- tem, and the extent to which that in- come could cover the costs of a highly- finished, durable and climatically appropriate solution to EWS housing.

Estimates for the finished houses, without return from product sales, ran to between Rs 55,000 and 60,000 which works out to approximately Rs 978 - 1,086 per sq m. Clearly, firing brick and tile as the only products could not generate enough cash. A line of terracotta household items (lamp bases, candle mud brick wall stands, garden stools and lanterns) was designed to meet the need for income generation. In addition, 250 smokeless chullas were fired to demonstrate the use of the house as kiln in making products for the village context. Though considerable difficulty was experienced through a summer of un- usually heavy rain, the project was ultimately completed with the following results:

 

Cost of 2 houses                   Rs 62,000

Cost of R+D              Rs 3,000

                                    -------------

                                    Rs.65,000

Less return on

product sales             -Rs.45,000

Cost of 2 houses

subsidized by product or,

about Rs.369 per sq mRs.20,000

                                    -------------


Fuel consumption in this system, which was fired as a longitudinal cross-draft (or zig-zag) kiln, was reduced to 6,800 MI per 1000 bricks.

The test demonstrates that considerable income can be generated in a small system This test is fully documented in Indian Architect but, quite frankly, not how that income and Builder (Nov 1990). The figures here are in general would really get back to the somewhat different as the IAB article went to press houses. 11 points towards the formation of small building cooperative of brick makers, potters, and masons, with overheads kept as low as possible. Even then, the cost per house is bound to be somewhat higher. That higher cost might be off-set in a situation where on-site clay is suitable for brick making. In this test it was not, and had to be transported two kilometers to the site.

This test is fully documented in Indian Architect and Builder (Nov 1990). The figures here are somewhat different as the IAB article went to press before the project was complete. Ray Meeker