Hand-thrown earth walls

Before we came to Hana, we knew it rained 80 or 90 inches a year, but we didn't know how often and for how long it rained on any given day. Having worked in Nicaragua, Brazil, and the cloud forests of Veracruz Mexico (not to mention northern California in the winter) we knew building rammed earth in the rainy season was challenging. Soil gets too wet to compact, tractors get stuck in the mud, and in general, wall quality suffers.

For these reasons, we made the decision to build at least the first of our structures under cover of roof. And since you can't ram a wall with rafters and sheeting overhead, we decided to use the hand-thrown method. Actually, I don't know of any other builders who do hand-thrown, but we've been experimenting with the technique off and on for a dozen years. Hand-thrown is a closet cousin of pise, our high-pressure air delivery method of building structural stabilized earth walls. Both techniques utilize a single sided form, a wet mixture impacted against the form and then shaved off to the desired wall thickness. In production pise, volumetric mixing machines and huge air compressors blast stabilized earth at a rate of 8 to 10 cubic yards an hour. We use the technique to build large structures fast. Hand-thrown, on the other hand, is slow, but effective; good for the pace of island building, and with cement stabilization, very resistant to weathering as we have observed over the last seven or eight years. (In fact, I was pressure-washing the house just today to clean off the salty residue that blows in off the ocean and the walls are like basalt.)

We built the walls in panels, corners with three-foot wings, and three-foot in fill panels between the big window walls. Each panel was reinforced with a grid of #4's at 12", with the bars threaded through holes drilled in the galvanized steel support posts. The wall formula was a combination of site soil, mined during the first phase of the excavation, sand and gravel imported from the isthmus near Kahului, and 10% Hawaiian Cement (imported from China). We measured proportions with five-gallon buckets (one sack of cement fills two buckets0, mixed with an Imer three cubic-foot drum mixer (brought over in the shipping container) and delivered in a wheelbarrow. The corners were about one cubic yard, the infill panels about a half yard.

Building a hand-thrown wall is pretty much like the name implies. Wearing gloves, grab a handful of the mix (it should be the consistency of paste) and throw it against the form. Start at the bottom corner and toss one handful after another against the form, working your way first across the bottom the full width of the panel. Keep throwing handfuls working across the panel, building both out to the finished thickness and up against the form. As you build the wall upward, keep the slope of the working surface at about 45 degrees. If the mix is the proper consistency, each handful will bond to the previous and the mud will quickly take an initial set. If the mix is too wet, too low in clay, or you manage to work too fast, you'll get some slumping. Try to avoid slumping.

After you've built the wall out to just beyond the finished thickness and up to about two feet in height, you'll want to shave back the excess to create a plumb, flat wall surface. We use a gunite finishers tool called a rod (or knife), but you can also use a 1x6 with a bevelled edge or a 24" concrete finishers trowel. The longer the cutting tool, the straighter the wall will be. Keep in mind, however, that the cut side of a hand-thrown wall is by nature an imprecise finish. Getting your timing right on cutting the wall will take a little practice.

Once the first "lift" is shaved, go back to throwing. When we start moving up the wall, we use plasterer's mortar tables to work off. Shovel the mud out of the wheelbarrow onto the table. It's much less work to grab a handful off of the table than out of the wheelbarrow. Some of us even prefer the two-hand toss, so you can get more material onto the wall at one time. You don't have to throw hard, just enough to get a good bond and embedment to the steel. Your arms get tired by the end of a panel. We tried to build one corner or two panels a day. Evan was here at the start, then Lucky (a local) did rest of the mixing, Khyber did most of the throwing, all of the cutting, and I helped out a little.

We used the hand-thrown method on only the first of the three main pavilions, switching to traditional rammed earth for the other two. The advantages to hand-thrown are that (1) you can work under a roof, (2) you only need one form face rather than two and a whole lot less bracing, and (3) the wall can be much thinner because the roof is supported on the embedded posts rather than on the wall itself. The disadvantages are that you'll need a higher cement ratio to make up for the lower compaction factor and of course you need to supply and set the posts.

The beauty of the shipwreck research center is that there can be no failures. Everything is a study in island appropriate methods and materials. We think up a technique or a combination of building materials, design a project to implement them, build it, evaluate the process, then watch it respond to the elements.

The photos in this post illustrate the building of the hand-thrown walls, plus I've included a photo of shooting pise on the mainland.

Framing with poles

One of the most challenging strategies to our shipwreck building research was to develop a roofing system that could either grow on the island or be simple, lightweight, and easy to import. Our first thought, of course, was bamboo with thatch. Bamboo is easy to grow in the tropics and sub tropics, is now being grown commercially on Maui and the big island, and Bamboo Technologies, headquartered on Maui, has won code approval for one species, bambusa stenostachya, grown in Southeast Asia.

Bamboo for construction has gained a lot of momentum in the past decade, thanks in large part to the work of architect and builder Simon Velez. I was pretty certain bamboo poles for roof structure, in combination with rammed earth walls, would be two key components on our ultimate building materials pallet.

Bamboo Technologies produces pre-cut kit homes using plywood and bamboo panels for both walls and roofs. I was hoping to get away from the plywood and use braced poles for the roof framing with a separate, non-structural roof cover. I still wasn't sure what the roof cover would be. I didn't think thatch would have the durability we were after; plus it's a little too much of a stretch even for my unconventional approach. Clearly we had some things to learn and experiments to conduct before we were ready for bamboo roofing.

As a starter, and to make obtaining our permit a little easier, we decided to introduce the building department to stabilized earth first and move into bamboo later. But, wanting to get some experience working with round structural members rather than sawn wood, we decided to use galvanized steel poles. You can buy them in bulk from fencing supply companies, and they are relatively inexpensive. We designed the roof with trusses and bolted connections similar to those we had seen in Simon Velez's book. It's all very light-weight and can go up quickly.

We had cut and drilled the members for the trusses before loading them in the shipping container. The first of our pavilions was going to have 6" thick walls of hand-thrown earth rather than traditional rammed earth, so we set two rows of poles in the foundation as we were pouring the slab to receive the trusses. If we had built 12" thick rammed earth walls, we could have skipped the poles, but since we didn't know how often it rained, we thought it would be best if we put a roof up quickly and then worked under the roof to build the earth walls. As you'll see in later posts, we did switch to traditional rammed earth when we found out that most of the time the rains are short and sweet.

As you can see in the photos included with this post, the roof framing did go up quickly, thanks to all the help from our friends who flew over the join the first of many work parties.

Getting started on site

Just as soon as we had the containers on the ground, Bullets and his dozer-driving partner Brandon graded the pad and dug the footings for the first structure - a central kitchen/dining/living pavilion. Since the building pads were on a slight slope, we were able to accumulate enough soil to use in a wall formulation comprised of 50% site soil, 25% imported gravel and 25% imported sand. I"ll cover the wall building for the first pavilion in a later post.

We had packed the container with 2x12 form boards (that we would later use for the rammed earth wall forms) and with the reinforcing steel for the footings and slab. The steps after cutting the pad were (1) setting the perimeter form, (2) tying the reinforcing and (3) installing the electrical service panel and conduits. The electrical supply came up from a temporary pole we had set along the access road. We poured the footings and slab two days after off-loading the containers. I had intended to color the slabs with a powdered dye dusted onto the fresh concrete and bull-floated in, but I hadn't counted on the long haul from the batch plant. When the mud arrived on the job it was just too close to set up that we couldn't get the dye to work it. In retrospect, we would have been better off bringing bulk sand and gravel onto the site and mixing our own concrete (which we did on subsequent pours).

The photos at the end of this post show one fun shot that illustrates the containers in relation to the first pavilion, the others are of the form boards and reinforcing steel, the service panel and conduits, and pouring the footings and slab. The photo at the top of the post is of Hamoa Beach.

Packing for the Trip

The adventure on Maui wasn't really like being shipwrecked, because we had plenty of time to plan what to take. Crusoe and the Robinsons had to forage from the wrecks for their building materials. We had months to complete drawings and compile extensive lists of what we wanted in the containers. The challenge we set for ourselves was to see if we could pack every single item we would need to build the first house and furnish it. We included a solar water heater, stove, refrigerator and even our old Kubota tractor. We were trying to prove that small shipping containers could be packed and converted into disaster relief housing. As we found out, one container would have worked, but the second got us started on the elaboration of our concept to include more local resources. Once we sourced materials available within our immediate vicinity the construction expanded to a laundry room, shop and several other auxiliary buildings adding, those elements to our building palette.

In January of 2002 the experiment began. We trucked the first two containers to the port of Oakland, then flew over to Maui to meet them. When we first purchased the land for our project, it was so overgrown with cane grass and wild guava trees you couldn't see the ocean, but we'd cleared a quarter of an acre where the first house was going to go and where the containers would be off-loaded. The idea was the containers, once emptied, would serve as temporary shelter and job trailer.

Getting the containers from the port in Kahului out to the building site was a challenge. For one, the road is narrow and winding. The bridges are one-lane and packed with a constant flow of tourists circumnavigating the island. The second, they were too heavy for a single forklift to off load them. The first problem we solved by hiring local hauler Lloyd Redo who makes the trip before dawn. The second problem we solved with 2 forklifts balancing each container delicately as they danced them into place.

Once the containers were on the ground, lined up precisely with mountain and ocean we were ready to start. The photos in this post show the containers loaded on the truck in Napa, being off-loaded, and our new friend Bullets, our backhoe driver. The first photo is a view of the island Alau wearing a rainbow.

Shipwreck - Building Rammed Earth on an Island

With both the Jack house and the Terra house winding down, I find myself with time to tell the story of building our family project in Hana, Maui. As a kid, I read and re-read Robinson Crusoe and Swiss Family Robinson, planning how I would build if I were shipwrecked on a tropical island. In a way, our Hana project is fulfillment of those childhood imaginings.

We started in 2002 with the first of three main pavilions, and each year we add to the compound: a guest bedroom, outbuildings, the master bedroom, the guest cottage, and the shop. At the time we started, I'm pretty sure it was the first rammed earth in Hawaii. Since then, Olson Kundig designed a really wonderful house at Slaughterhouse Beach, and we're starting to consult on a project in Hawi on the Big Island.

I believed then and I'm convinced now that rammed earth is an ideal material for island building: it replaces imported materials with native, is non-toxic, eliminates wall cavities where insects and rodents congregate, it holds up to the salt air, wind, and rain, and it doesn't need painting. The current common building practice in Hawaii is much like mainland building: wood frame, exterior siding, fiberglass insulation, gyp board and paint. All the wood is imported and chemically treated, and the houses are built tight to keep out the bugs, which then traps moisture and fosters mold growth.

Traditional Hawaiian housing was made up of a broad roof on poles covering an open living space. Walls, where they existed were woven and allowed the wind to blow through. For our project, we wanted to develop an architectural style and a pallet of materials that respected local tradition and that would acknowledge vernacular successes. We designed our buildings with broad roofs on poles, open living spaces and moveable walls. We intended the wind to blow through to keep the house comfortable naturally. Mostly we wanted to build affordably, using local materials whenever possible and as simply as we could.

We're in Hana now and for most of the winter. Over the next couple of months, I thought it might be worthwhile to go through the digital scrapbook and put up some illustrated posts of the processes we've developed appropriate to the region and the resources. You can extrapolate them to your own projects. Put your comments on the blog and I'll respond if I can. The photo at the top of this post shows two of the containers converted to a cottage.