Tuesday, September 07, 2010

Rammed Earth Home Construction in Chad

This post, the third in the series by Mark David Heath, will offer an update on the previous two projects as well as where we seem to be headed with rammed earth in Chad.

Our first project of rammed earth in Chad was the wall around our work base in the village of Mainani, near Kome, the base of ExxonMobil in Southern Chad. We built this first wall, 676 meters long (2217 feet) by 1.5 meters tall (4ʼ-11”). This site was originally used as our work base for the road work we were doing for ExxonMobil. However, with that work concluded and our work now being the composting of ExxonMobil waste material, we needed to raise the wall height. The first wall was built of laterite soil, taken straight out of the borrow pit, and screened to remove pieces larger than about 36 mm (1.5”), mixed with 5% Portland cement.

With the added height to the wall (we added about 800 mm (2ʼ-7”)) we used laterite soils again, but this time, we used “recovered” laterite soils from the roadway windrows. Through the normal course of maintaining an earthen road, soil is normally cut off the road and pushed off to the sides, creating the “windrow” typically seen all along the sides of earthen roads. If these windrows get too large in areas of high rains (like Chad) they can become a drainage problem. In a particular section of road we were building, these windrows needed to be removed to relieve a drainage issue. Since this material is typically being graded off the top of the road, it has had a lot of traffic wear and tear on it. It has been crushed and ground down by the traffic, resulting in a much higher percentage of “fines”, material that is much less coarse than the laterite material coming right out of the borrow pits.

This laterite material with higher fines in it provided results more like what we had experienced with the clay soils and sand. The higher percentage of finer material produced a much smoother and more tightly consolidated surface that laterite straight out of the pit. We still added 5% Portland cement to the majority of the extended height sections. We set our wooden forms right on top of the existing rammed earth wall, now over 4 years old and facing the fourth rainy season.

We also did an experiment with “non-stabilized” rammed earth. All of our previous projects had used some percentage of Portland cement as a “stabilizer” to make the finished product more weather resistant. The biggest issue for long-term use of rammed earth is weathering, especially due to water, in any form - rain, snow, ice, etc. With Chad having no frost, this is to our advantage. However, with all of Chad having over 30” of rain per year, rain damage is a huge issue for rammed earth in Chad.

However, I had read of structures being built of “non-stabilized” rammed earth, meaning without Portland cement, lime, asphalt emulsion, etc, as a stabilizer. I have also seen rammed earth structures several hundred years old, built with non-stabilized rammed earth. Given the possibility, and given that the expense of the Portland cement is our largest single expense in the rammed earthh, we decided to give non-stabilized rammed earth a test.

The issue in non-stabilized rammed earth is “getting the mix right”, meaning getting the right combination of sand and clay to produce a “stable” and durable structure. Fortunately, I was shown an old and simple test for determining the “right mix” which we have used very successfully. We took the principle of the test and then adapted it to our circumstances, here in Chad. Since we already have forms built for making corners of 500 mm square, and since all of our materials are mixed by hand we made up seven test blocks, using our 500 mm x 500 mm forms. We mixed up 30 shovels of material, in varying quantities, and then rammed a block with each different mix. We chose 30 shovels of material per block because it is relatively simple math to make the percentage calcs with, and a “Chadian” wheelbarrow will hold just about 30 shovels of material. Our blocks were as follows:

Clay soilsSand%Clay Soil%Sand
18 12 60.0% 40.0%
21 9 70.0% 30.0%
24 6 80.0% 20.0%
Across seven test blocks, we are able to look at clay:sand ratios varying from 20%:80% to 80%:20%, with each block having 10% more of one material and 10% less of the other material than the last block. Our objective is to see which mix is the most durable. We let them dry for at least 3 days and then, if there is no rain, we “make it rain” by watering the blocks with a garden watering can, the same one we use to add water to the rammed earth as we are mixing it for ramming. We dump about 20 Liters (5 gallons) at a time on each block, and do it 2 or three times a day.

After about a week to 10 days of “rain” we look at which test block held up the best. In Mainani, with our “recovered” laterite and sand mix, test block #5 held up the best - made up of 18 shovels of laterite and 12 shovels of sand. Since this was all new to us we were a bit cautious. We built the bulk of the wall with “recovered” laterite and 5% Portland cement, like the mix we had used to build the original rammed earth wall, 4 years ago. But, at the back of the property, where no one from the street could see (just in case it was a disaster) we built a part of the new wall with non-stabilized rammed earth, using the mixture of test block #5. We found out, very quickly, that non-stabilized is much more moisture sensitive in the early days, as it is drying, than is stabilized rammed earth. A couple of times we had rains hit us just as we were finishing or that same night,  and we found our non-stabilized rammed earth wall, living up to the name “non-stabilized”, lying on the ground when we came back in the morning. With that lesson earned, we bought some plastic sheeting and covered the fresh walls for at least three days.

Right after completing the non-stabilized rammed earth sections we had some very strong rains, including one that lasted almost 18 hours, and a couple that, while not a long, were even more intense. We were more than pleasantly surprised to see that after 2 weeks of heavy rains, preceded by at least 3 days of good drying of the non-stabilized rammed earth, that there is virtually zero difference in the non-stabilized rammed earth vs the stabilized rammed earth.

This knowledge, combined with the knowledge gained from our second project (described next) is providing what we think is the direction to now be headed with rammed earth in Chad.

In Doba we did a joint venture project of rammed earth on a residential lot. The owner put up the lot and we did the rammed earth. We have recently sold this project, recovering our investment, the land owner recovering hers, and we both made a small profit.

This project was on a residential lot measuring 25 m by 40 m and is a ʻcorner lotʼ, having streets on two sides. We decided with the owner to put two gates on the long side street. Each gate opening is 3 m wide. Unlike the Mainani site, we are trying to do a demonstration of a more typical rammed earth project. So, we have a footing under the rammed earth. We chose a gravel-filled ditch for our foundation, similar to the idea of a railroad foundation. Gravel costs less than concrete, does not need Portland cement, and distributes the loads well and allows water to pass freely, which is a concern here where we get more than 1 000 mm (39”) of rain per year. So our foundation is a gravel-filled trench 500 mm wide, which is twice as wide as the wall, and 300 mm deep, which gets us past the topsoil and into the subsoils.

We began with the installation of the corner posts which were made with 5% Portland cement. Then we installed the remainder of the intermediate posts, again with a 5% Portland cement mix. After all the posts were built, we built the wall sections, between the existing posts. For the Wall sections, we used a mix of 2% Portland Cement.

As you can see in the photos, all of the posts are “keyed” so that the Wall sections tie into them with the keyway. The intermediate posts are “Tʼs” and we built an intermediatepost at least every three wall sections. The wall sections are about 3 m (10ʼ) long by 2.4 m (8ʼ) tall. This project was covered in greater detail in the previous article.

After having built the walls and we build two small buildings inside the walls. One was a one-room structure that could serve as a guard house, situated between the two gate openings. The second structure was a 60 m2 structure with a bedroom, a bathroom and a “great room”, a living room / kitchen combination room. The photos show the windows formed out and then having a beam poured over them as we poured the bond beam.

This project received a lot of interest and a lot of folks coming to look at it. They spoke well of it, of how thick the walls were, etc., but we had no one willing to buy it. Our partner was thinking that she was going to have to live in it to prove to the Chadians that it was a good product. Then, one day, we were talking about the project and she said that she has overheard someone talking about how they did not trust the project because the rain had eroded the walls. We had built this right at the beginning of the rainy season and we intentionally did not cover the walls to see just how much damage the rain would do to fresh walls. What was a good thing to us, showing what we felt was very little degradation, seemed to have become a really big negative to anyone who came to see the project. The owner suggested that we plaster the walls. This would cover up the rain erosion and it would make the walls look more “familiar” to the Chadians.

Traditional brick walls need to be plastered or the rains will wash away both the bricks and the mortar as can be seen. So, even though we understood that the minor erosion that had happened to the Rammed Earth wall while it was still fresh and before it had hardened was all that was going to happen, the Chadian thought that it would continue and eventually eat out the entire wall, as happens with traditional bricks.

So we plastered the two walls facing the street, being the two walls facing the prevailing wind, which were the two that had taken the most abuse and suffered the most erosion. We plastered them with a “tossed” or “thrown” plaster made of sand and Portland cement.

The result was that the walls looked just like any other plastered wall, except for being twice as thick, and, that they were immediately accepted by the Chadians. By the time we had the first two faces plastered we had 4 offers to buy the project, and before we started on the third side, we had it sold, and cash in hand. Another demonstration that, especially with new products, you need to be sensitive to the customer and find a way to get them to accept the new product.

With the outcome of these two projects, the one with successful non-stabilized rammed earth seems to be offering us a more clear direction in which to proceed with rammed earth in Chad. Since the Chadians prefer the stucco look, and since non-stabilized rammed earth can be built to deliver similar performance to stabilized rammed earth, then it seems to us that the way to go is to build  non-stabilized rammed earth, then plaster the walls. This save the expense of mixing Portland cement in the rammed earth, delivers the same performance, as long as the rammed earth mix is correct, and the saved Portland cement can be used to plaster the walls, making them acceptable in the marketplace and even more resistant to the rains.

This new direction could really help in the general acceptance and diffusion of rammed earth technology in Chad. Rammed earth is faster than either CMU (cement block) construction or traditional brick construction. Rammed earth is much thicker than traditional brick and stronger than CMU or brick. This offers greater security to the occupants. Being thicker, the heat of the sun will not get through the walls, making the structures built with rammed earth more comfortable. Now, the knowledge that  non-stabilized rammed earth can be built with comparable performance to stabilized rammed earth, means that rammed earth is now economically competitive as well. It was already less expensive than CMU construction but more expensive than traditional brick. Now we are even less expensive than traditional brick construction, as well.

The single biggest roadblock seems to be the initial cost of the forms. Although this is not a huge issue for an existing construction company that is going to be going into the rammed earth business, it could be for a small enterprise. We are looking at trying out a forming system that will make large “blocks” of Rammed Earth, about 1 000 mm long, 250 mm thick and 500 mm tall. These would have to be built in a “bond” pattern, similar to laying CMU block or brick, but this technique is already well understood. The initial cost for these smaller forms would be significantly less, and, with two sets, the speed should not be too compromised. We still need to sort out how to build corners, and intermediate posts.

The owner of the Doba project is now so pleased that she has formed a small construction firm, hired a crew, and is building a project for a new client and has already signed two more clients. We also have a young, newly graduated engineer, whose father has a long-established construction company, doing an “apprenticeship” on these first new projects. So, we have what looks like the first two adoptees of the rammed earth technology in Chad and we have the third project underway, and two more right behind them. The first project, in Mainani, is being actively used as a yard for making compost from Esso waste, the second project, in Doba, has been sold and is nearly completed and ready to be occupied by the new owner. The third project is underway and going well with a new adoptee and a new apprentice. Rammed earth in Chad may be off to a new role in the battle against desertification.

When we first built the walls in the second project we thought that we needed to have at least 2.5 to 3% Portland Cement in our mix. We know now that we were basing this on short-term issues. With higher cement content it is true that the walls eroded less in the rain. However, with the correct mix, we now know that the walls get very, very hard with time and if we protect them from early rains, we can build excellent structures with 2% Portland cement and without any Portland cement, at all.

What our experience has now shown us is that rammed earth takes a significant amount of time to fully harden. However, with that time, it gets really, really hard. When we went back to cut in the windows on the walls where we did not form them in, we found that our rammed earth walls were every bit as hard and as difficult to break open as the concrete construction in the area. The crew assigned to cut out the window openings kept going on about how surprisingly hard the walls were. There were a couple of doubting Thomases that were converted watching the windows being cut into the rammed earth walls that had now hardened for one year.

1 comment:

aman said...

I like the way you start and then conclude your thoughts. Thanks for this nice information regarding regarding I really appreciate your work, keep it up. ecofriendly construction