Sunday, June 07, 2009

Rammed Earth In Chad: The Sequel

This article will offer information on a follow‐on demonstration of Rammed Earth in Chad, in a project carried out in Doba, a small town in southern Chad, Africa, about an hour northeast of Mondou, and on the edge of the ExxonMobil oilfield. The first document, prepared on the first demonstration, had quite a bit of background information in it, which this document will hopefully not unduly repeat.

The first demonstration was carried out in a village called Mainani, near Kome, the center of the ExxonMobil oilfield in southern Chad, Africa. The first demonstration was a wall made of laterite soils (laterite, noun, a reddish clayey material, hard when dry, forming a topsoil in some tropical or subtropical regions and sometimes used for building. Geology, a clayey soil horizon rich in iron and aluminum oxides, formed by weathering of igneous rocks in moist warm climates. From my MacBook Dictionary). We added 5% Portland cement to the laterite soils. The forms were made from waste steel shipping boxes from ExxonMobil and the ramming was all done by hand, by local village women, using their traditional pestle sticks which are used daily for grinding flour from millet and rice.

In the article on the first demonstration, we described the forms, the mix, the
ramming, and the results. We also discussed what we thought would be our next steps in further demonstrating the Rammed Earth technology in Chad. Here we will begin with some brief updates on the first demonstration.

It is now about 3 years since the first demonstration was carried out. The first wall has been through 3 full rainy seasons and the 4th has just now begun. The first rainy season was the heaviest in the previous 10 years and caused heavy flooding and there were many traditional buildings that failed and collapsed. The last rainy season, made the first look mild. It was reputed to be the heaviest rain is over 25 and possibly 50 years. The flooding was far more extensive than 2006. Many traditional houses collapsed, roads were impassible and washed out, and many crops were destroyed. The first demonstration wall, made of Laterite soils and Portland cement, has well withstood the past three years of very heavy rains.

The wall is very slightly weathered on top and on the face. Much of the weathering is due to the nature of the laterite material, itself. It is quite gravely. The final surface of the first wall was quite rough, right from the beginning. The ensuing years of weathering have only slightly increased the roughness of the surface. Also, the wall was built “doing everything wrong”. We intentionally wanted to test the limits of the wall to face the likely mistakes of future users. For example, we built the 200mm (8”) thick wall, about 2 meters (6’) tall without any foundation under it. We built it right on top of the uncompacted existing topsoil with 200mm scrap steel posts from ExxonMobil set 1 meter (3’) deep at 6 meters (20’) on center. That is a pretty poor way to build a heavy earthen wall. But, again, the objective was to make as many ‘mistakes’ as we thought we might have future users make. The result is that there has been some settlement of the topsoil in some portions of the wall causing some minor cracks. None of the cracks has opened enough to fit a playing card in it. But most of the sections are uncracked and looking very much like the day we pulled the forms off.

In this new demonstration we wanted to accomplish several things:

1. We wanted to use clay soils and sand, rather than laterite soils. Laterite soils are readily available in the area of southern Chad, but not so elsewhere. On the other hand, clay soils and sand are available most anywhere in the country, and is the most common building material used in Chad.
The conventional method of construction in Chad is to make bricks of clay soils, dry them in the sun for a couple of weeks, stack them and bake them with wood or charcoal fires, then construct the structure using clay and Portland cement mortar. Then, the norm is to plaster the surface with Portland cement and sand plaster. If the brick structure is not plastered, the mortar and the bricks can suffer badly from the rain and weather, and it is common to see bricks starting to fall out of walls after 1 to 2 years, and commonly, by the third year, there will be more than a few ‘gaps’ in the wall. So we wanted to use the common materials of clay soils, sand and some Portland cement, but in a “new” and better way.

2. We wanted to use a forming system that could be readily duplicated anywhere in Chad. Our forms for the first wall were made of scrap steel from ExxonMobil. This time we made them from the rough‐sawn wood that is common in the Chadian marketplace. We began with the general concept presented by David Easton is his pamphlet and later book on Rammed Earth, by the same name, and modified it to accommodate the lumber available to us here. David used 1‐1/8” thick plywood. There is no such thing here. So we used 30mm (1‐1/8”) thick planks about 300mm (12”) wide. Because the lumber is rough‐sawn, we hand‐planed the planks down to a uniform 290mm width. These planks are the faces of the forms. The planks are backed up by 40mm (1‐1/2”) thick by 80mm (3”) wide members at about 150mm (6”) on center. Horizontal members, made of the same planks used for the form face, but doubled up and nailed together, and set on edge, then back up these vertical members. These horizontal members are held in place with 3⁄4” diameter pipe clamps. The pipe clamps are not commonly available in the marketplace but they are not difficult to import. All the alternative materials we looked at in the marketplace were much more expensive and not as simple and flexible in use. All of this is shown in the photos.

3. We wanted to go higher and not have to use scaffolding. The formwork is strong enough that the forms can be stacked over 3meters (10’) tall and the workers mount and work on the formwork, itself, as the ‘scaffolding’.
Granted, it will give a Cal‐OSHA and ExxonMobil Safety Inspector pause, but I climb them daily and the workers use them without any trouble and we have had no problems and no ‘Near Miss’ incidents nor “accidents”.

4. We wanted to continue to use the village women to ram the earth. Writing of the women ramming right after the above note on safety reminded me of an ExxonMobil Safety Inspector coming out to see the first demonstration wall. The village women were ramming the wall in native dress and either barefoot or with ‘flip‐flops’ on. He commented that not wearing safety shoes was a hazard and suggested we give the women safety shoes. Not wanting to get into a conflict on such a minor matter compared to the bigger picture of presenting a new system to the local market, we agreed, and gave the women safety boots. Within two days the women revolted. They had never worn shoes in their lives. The boots were chafing their feet, they were hot and their feet were sweating and beginning to have damping problems. Also, the weight of the shoes was putting them off balance, and they felt insecure on the scaffolding wearing the safety shoes. I asked the women how many of them had ever hit themselves with a pestle stick and they responded by laughing at such a dumb question. I asked if they had ever hit another person, and one women asked “you mean accidently or on purpose?” So, we asked the women if they would like some tennis shoes instead of the safety boots, and they were 100% in agreement. We decided that the biggest risk was not the pestle sticks but bare feet on metal scaffolding, so tennis shoes offered the needed protection. The women finished the wall, in tennis shoes, happy and accident‐free.

5. We wanted to make sure we were more environmentally “appropriate”, more “friendly”, than traditional bricks. The traditional construction with clay bricks requires the destruction of a large number of trees in order to bake the bricks. It is certainly more appropriate to burn a tree that took 25 years to grow in order to make bricks that will last at least 25 years. However, if we can make an equally durable structure without cutting any trees, then we are ahead of the game. In Chad, desertification is the #1 environmental issue and the desert is moving south at shocking speed. In the few years I have been here I can clearly see the advance south. Also, traditional construction requires that the bricks be plastered. Otherwise, the mortar and even the bricks will not last long in the rains and weather. This plastering can be mud plaster and then re‐done each year, after the rains. Or it can be done with Portland cement plaster and last for many years. The amount of Portland cement needed to plaster the bricks is far more than the amount we use in Rammed Earth construction. Last, traditional construction builds with a brick that is approximately 90x140x250mm (3.5”x5.5”x10”) laid in a “common bond” pattern (one brick, centered on top of two, laid length‐wise). (As an aside, I have yet to figure out where this brick size came from, because it is not modular. You cannot lay them in a double wall or corner, etc, and “make bond”. A modular brick is twice as long as it is wide, including mortar joints, so that it will lay out correctly at corners, etc.) When traditional walls are constructed one brick thick, one wythe, they are miserably hot structures. The house really serves as a secure place to keep things and a place to keep the rain and sun off. The people only sleep and work inside during the rainy season, when the sun is greatly blocked by clouds and the temperature drops. Otherwise, the people sleep outside, most of the year. I have made many ‘tests’ of traditional brick walls where I go to the outside of the wall at the end of the day and feel the temperature of the wall, and it is hot to the touch. Then I go inside and feel the wall surface and it is just as hot, if not often hotter than outside. With our rammed earth construction we are making the walls much thicker and the sun is not able to well penetrate the thickness of the wall. Every time I have checked our walls, the side facing the sun is hot to the touch, while the shady side is always clearly cool to the touch. To accomplish the same thing with traditional brick would require at least twice as many brick and the consequential destruction of twice as many trees, not to mention being twice the cost and time.

6. We wanted to be economically competitive. We have done several cost comparisons and had our numbers looked at by a couple of local construction firms. It looks like our rammed earth walls are about 30% less costly than
local conventional brick walls and we are about 50% faster. I have just heard that we may be able to get lime here in Chad. I have tried to find lime but the imported bags of lime cost more than Portland cement. But I was just told that north of N’Djamena, the capital, there are very large deposits of lime. I was told that the lime is so abundant that the price is the same or lower than sand. This could really be a great improvement because lime can be used instead of Portland cement and with excellent results. If it is true that we can get access to a Chadian supply of lime, and at a lower cost than the imported Portland cement, that would help both the economics of Rammed Earth as well as make it environmentally more appropriate with a locally produced source instead of an imported one. Currently, traditional fired‐brick walls cost about 13 000 XAF/square meter ($2.40/SF), without plastering the faces, and our Rammed Earth walls are running about 8 000 XAF/square meter ($1.50/SF), and do not need to be plastered. We are also trying to pay a decent wage. The national average wage in Chad is about $1/day and we are paying between $3 and $4 per day. I might also note that, at that wage, we have people coming and standing on the edge of the jobsite looking for work each day. Sometimes as many as 20+ people are asking for work. Such is the life in the 4th or 5th poorest country on the face of the earth, and 40 to 70% unemployment (depending on whose stats you like most). In this demonstration with clay soils and sand we began by testing the mix of clay soils and sand to find the optimal ratios. We made 7 samples with Clay Soils: Sand ratios of 4:6, 5:6, 6:6, 7:6, 8:6, 9:6, 10:6. All measurements were made with ‘full’ shovels; since that is the way we were going to be measuring the mix for the
production of the wall. Our best mix was the 9:6 mix which is 60% Clay Soils and
40% Sand. For those of you with some knowledge of Rammed Earth, you may
immediately be thinking that 60% clay is far too much clay, because the traditional mix is about 30% clay. My explanation is that we are using “clay soils” and not pure clay. The clay soils are very sandy so, in truth, we are hitting about the 30% pure clay content.

After finding our best mix of Clay Soils and Sand we then made 5 samples with Portland cement to determine how little cement we could use and still get excellent results. Our samples were 5%, 2.5%, 2%, 1.5%, and 0.5% of Portland cement, measured by volume. The 5% mix was like concrete. The 1.5% and 0.5% mixes were softer than I was comfortable with. I am sure that at the end of the full 28 days curing period, which we really ought to be giving the Portland cement, that all the samples will be just fine, but at the end of one week, which is when we made our choice, we decided to not go below the 2% ratio.

Our next decision was kindred to the decision with the first demonstration with laterite soils: do it “all wrong” and see how it held up. Since this demonstration is being done in a new ‘subdivision’ area of Doba where many families are moving to, we wanted it to be good enough that it will not be a risk nor an eyesore to the neighborhood, and maybe even a finished product that we would sell. So, we were not ‘as crazy’ as the first time. This time we started with digging a footing 450mm (18”) wide by 300mm (12”) deep. We filled this with compacted gravel. Then we build the property wall right on top of the gravel. Because of the mass of the wall, about 2 000kg/cubic meter (140 lbs/CF), and the fact that we are in a non‐seismic area, the wall really only has to resist wind loads, and the mass of the wall, 250mm thick (10”) and sitting on compacted gravel is enough to resist the wind loads.

We decided to make the ‘corner posts’ and ‘interim posts’ of the 5% mix and the wall of the 2% mix. When we built the corners and interim posts, we did not bury them, as we normally would have done. We just used the stronger mix and built them right on top of the compacted gravel footing. Each post and wall segment is ‘keyed into’ the next with a 70mm (2.5”) by 30mm (1.25”) key. The wall segments are about 3meters long (10’) and we are putting a post every 4th wall segment. The corners are 500mm (20”) on each side and the interim posts are “T” shapes 750mm (30”) wide by 500mm (20”) deep. There have been lots of folks coming by to hit and push on the walls and posts and they are all still standing with lots of new found ‘believers’ walking away shaking their heads.

When we began, we first built the four corner posts. The demonstration is the property wall on a lot measuring 25meters (82’) by 40meters (131’). This is 1000 square meters, which is the rough equivalent of a 1⁄4 acre lot. The wall is 2.6 meters (8’‐6”) tall. This it ‘tall enough’ that the thieves cannot easily jump the wall, which is a constant threat here. The wall is also thick enough, dense enough, and strong enough to stop small arms fire, which is always on the minds of the Chadians; 16 years of civil war leaves a very deep, ugly, memory. As soon as the corners were built we built a wall section on each side of each corner post. Then we laid out the interim ‘T’ post on the short side wall and the corner posts for the 3 meter (10’) wide entrance gate opening. After that, we just began filling in the remaining sections of the walls. At this writing we are 2 weeks into the project and over 50% complete. We have a 3‐man formwork crew, 6 women ramming with pestle sticks, and 6 men mixing the soils by hand with shovels.

The rainy season is just beginning and it has been very insightful for the demonstration. We finished the first wall section on the day of the first rain. The rain came just after nightfall and there were a bunch of folks on the jobsite before first daylight. There were many comments about not sleeping well, listening to the heavy rainfall (about 3” per hour for 2 hours) and worrying that “their new wall” would be washed away. Again, in the spirit of doing it all wrong, I made sure that we stripped the formwork before we left and did not cover the wall. There were a lot of smiles the next morning seeing the wall section and all the corners standing straight and tall. At sunrise, about an hour later, there were lots more smiles and back‐patting and hand‐shaking when the first light of the day showed that there was only the slightest erosion on the windward side of the new wall, less than 12 hours old, and nothing on the corner posts, 2 to 6 days old. It rains every 2nd or 3rd day, with each rain lasting from 1 to 3 hours and 2 to 4 inches per hour, in intensity. When the rainy season is fully upon us it will rain every day, and can rain up to 12 hours, or more, and at the same intensity. Just as with the first demonstration with
laterite soils there were many saying that the wall would come down with the first rain. We have gained enough respect that we have already received 3 requests for quotations to build them a wall, and 2 requests for houses build of Rammed Earth.

When we build for clients we will be much more conservative. We will likely add 3 or 4 vertical rebars in the posts and a horizontal rebar at about 300mm (12”) on center, connecting the vertical bars. We will also bury the posts about 750mm (30”) in the ground and will likely put a 50mm (2”) mud slab under the gravel footing to ensure that the gravel does not get ‘pushed’ into the surrounding soils with the heavy rains. (Geotextile cloth could serve the same purpose as the mud slab, but they are not readily available here.) We will likely not make any changes to the wall sections, except for the mud slab under the gravel.

Once the wall is finished we are planning to build a couple of “model homes”. The first will be two single‐story, one‐room structures, similar to the most modest of the village homes. One will be square and one octagonal. The traditional house in the Chadian village is round for the poorest and a square house shows you have a bit more wealth. Each room added is a sign of increasing wealth. The traditional ‘middle class’ house is 3 rooms – the ‘Salon’ (living room), and two ‘Chambres’ (bedrooms). And this will be the third model. The bath area is usually an outhouse outside. Having indoor plumbing is solidly upper middle‐class and is never seen in the villages, only in the towns and cities. We will add this in the fourth model, which will be 2‐story and octangular. We might add a small indoor bath in the third model, plus a traditional outdoor outhouse.

Round construction is the tradition for reasons not clearly understood by any villager I have yet spoken to. Every time I ask why the poorest build small round houses, all they say is because they are poor. What seems to have gotten lost in the generations is that round buildings are more efficient, in several ways. It terms of labor and material expense, a square building will require 11% more bricks to make the exterior walls than a round building of the same floor area and same wall height. The villagers have lost this knowledge. Since the round buildings they build are typically much smaller than the square ones, in their mind, round is just for smaller houses. When they get some more affluence,they immediately seek out to mimic the European, “white man”, rectilinear style.

Round buildings are also more efficient structurally. They better resist wind loadings and are simpler to roof than a rectilinear structure. They also do not expose and entire flat face to the sun to be heated up all day long, and end up being more comfortable to live in. Here in a hot, sunny climate, having less wall surface exposed to the sun to collect heat makes the building more livable.

I did find one excellent example of modern construction in the capital where the architect, a Swiss‐trained architect, came to Chad and studied the local architecture and then designed an orphanage, respecting the traditional lines, but used pre‐cast concrete panels in an octangular structure and then put a brick dome on top and protected the dome from the sun with a grass roof. The buildings are surprisingly cool inside, without any air conditioning, and the design is very pleasing to the eye.

Since octangular structures are very similar to round structures, in terms of efficiency, we will be building an octangular model. Formwork for a round building is very difficult to construct, so making a circular rammed earth structure is not very efficient. But, making an octangular structure is not too much more complex than a square one. The angles require a bit more effort and skill, but the wall segments are all formed with straight panels. A square building uses 9% more bricks to make the exterior walls than an octangular building of the same floor area. So, the difference between round and octangular buildings is only 2%, while square buildings have about 10% more exterior wall surface than either round or octangular buildings.

Building with rammed earth is very kindred to building with formed concrete. In formed concrete I am constantly aware of my first job out of college where I was working for a major US builder who self‐performed their own formed concrete work. It was my first job to work on the concrete estimating team where I quickly learned that the formwork was the most expensive part of concrete construction.

With rammed earth this is also very true. The most expensive part of the work is the formwork. So, making excellent forms that can be reused many, many times is worth the investment. Also, using a system that is flexible, so that it can accommodate a variety of floor plans, is of great value. And, last, making sure that anyone designing with rammed earth understands that the formwork is the most expensive part of the work, will help them focus on designs in a ‘modular’ manner that allows for re‐use of the same formwork and does not require the fabrication of one‐off forms.

In our demonstration, the cost to fabricate the formwork is more that any other cost – the labor to ram and the materials, sand, clay soils, Portland cement. In fact it is almost equal to all of the other costs, together. But, the forms are designed to be reused at least 100 times, so the cost per use, comes way down. To make the walls, we have 4 “sets” of wall forms. The walls are being built ‘3‐forms high’ and this allows us to set up the 4th set while the top of the last wall section is being filled. As soon as the top of the wall is filled, the crew can immediately move to the first set of forms, already in place, on the next section of wall. While the mixing and ramming crews are filling this first set of forms, the formwork crew can strip the forms off the last section, making them available for the new section and one set to move on to the next section. This allows the work to move forward continuously, and to not have to stop at the top of each wall section while the forms are stripped, oiled and

We are very pleased with the mix of Clay Soils and Sand vs. the Laterite Soils. The clay/sand mix is a finer mix and so the surface of the finished wall is smoother and is fine enough to reveal the texture of the wood formwork in the finished wall. Many of the people who have been to see the demonstration have commented on the natural beauty of the mass of the rammed earth wall with the texture of the wood formwork left in the face of the wall. They have said that the surface is an ‘artistic’ surface and would not want it covered up with plaster. Since the surface is durable enough to resist the weather, there is no structural reason to plaster over the wall, unlike the traditional brick walls, which must be plastered over in order to survive. We will have to come up with some method for sealing the surface, especially inside the houses, since rammed earth, like concrete, is always ‘dusting’ at the surface, if it is not sealed. The challenge will be to find something that is already here in the marketplace and does not violate the economics or environmental appropriateness of the rammed earth.

In conclusion, the second demonstration wall is better than the first and the methods, tools, techniques and materials are all readily duplicable throughout Chad.

As soon as the wall is completed, we will move on to construct two small “village” homes and, after that, two “town” or “city” homes. With the excellent reception, the excellent results, and the proven economics of the product, Rammed Earth has a real chance to make a significant and positive impact in Chad. On a closing economic note, the Chadian government is trying to carry out a program of ‘social housing’ and trying to hit a construction budget of $15/SF. So far, the best that the market has delivered is about $25/SF. We have taken the results of the second demonstration wall and done some preliminary estimating and it appears that we will be able to meet the desired $15/SF target. The next edition will hopefully provide positive evidence of such affordability.


Mundo Novo said...

Looks great. I like your forming system. What does your form for the corners look like. I'm doing a house in Mexico and scratching my head at how to do the corners. Cheers

John Francis said...

I have been chewing on executing a similar project in Kenya. I would appreciate striking up a dialogue. I like your approach. It resonates with me. Thanks,

Michael Phillips

John Francis said...

I would like to execute a similar project in Kenya. I like your approach and would like to begin a dialogue. Thanks,

M. Scott Phillips

Rammed Earth said...

Michael, send your email to (at) and I will forward your information to Mr. Heath in Chad. Keep me posted on your project.


Unknown said...

hello M scott - I too am in Kenya preparing a farmhouse extension project in western foothills of elgon. we have sandy soil with pockets of clay, and Gneiss (i think) igneous/ metamorphic stone readily available. you requested dialogue: am ready - wordpress | coxfolio

Unknown said...

hello M scott - I too am in Kenya preparing a farmhouse extension project in western foothills of elgon. we have sandy soil with pockets of clay, and Gneiss (i think) igneous/ metamorphic stone readily available. you requested dialogue: am ready - wordpress | coxfolio

Unknown said...

having read to the end of the chad sequel blog - just to commend and appreciate your excellent documentation. however, a typical section drawing and method statement would be very welcome. rgds architect cox