Timber-frame-and-scrapwood Greenhouse

For years as a farm manager of an organic Baltimore-area CSA I watched in frustration as tomatoes and cucumbers became more and more susceptible to blights, molds, and rots, seemingly no matter what I did, and listened to our CSA members’ complaints when the fruits came in too late for their expectations. The final straw was when our 2009 strawberry crop got wiped out overnight by gray mold after several days’ soaking rain. There’s got to be a better way than this, I thought. And I knew there was—greenhouses with roll-up sides, or “high tunnels,” as they are known. 
 
The only question was, where to put them? And with what money? And how to build them so as not to collapse in a snowfall like that the Baltimore area experienced in February 2010, two back-to-back snowstorms with accumulations up to around four feet?    Ordinary pipe-frame hoophouses and greenhouses collapsed all over the countryside. And high winds have been whipping the area with ever-increasing frequency as well. How to build something that sturdy on basically no money, and very limited construction knowledge and experience?
 
So it was that in early 2010, I approached Heathcote Community in Freeland, MD, with the idea of constructing first one, then two or three timber-frame high tunnels using mainly on-site materials, as part of a way of increasing food production for that community and generating income for the grower. Heathcote has traditionally grown some food for itself from a series of small but well-managed gardens in its few open spaces; nearly all the property is wooded or wetland. My overall plan was to utilize some of the forest area for shiitake mushroom production using on-site oak logs and sowing ginseng on north-facing hillsides, then intensifying fruit and vegetable production through the construction of large high tunnels. Heathcote had already installed a fine 20’x24’ hoophouse on its only piece of open flat land that was not seasonally underwater. Nice start; I just needed to expand the scale about tenfold.
 
Heathcote is a community which is based on permaculture principles—an approach which emphasizes turning a site’s resources and its limitations to advantage, using imaginative design and careful siting as its most important tools. A recently constructed strawbale house with a huge roof catchment area lies about 350’ upslope from the site I eventually selected for the greenhouse, perfect for setting up a gravity- and rain-fed irrigation system. Runoff from the greenhouse(s) can then be used for residential use and irrigation for what lies downhill. 
 
The site that appealed to me lay just uphill from their existing “hill garden,” a clear space on a southeast-facing hillside near the community center. The site was under very recent (<20 years old), not-very dense tertiary-growth forest, full of multiflora rose, Norway maples, some young walnuts, black locusts, and wild cherries, as well as the shaded remnants of an orchard planted in the 1960s. I cleared about a third of a an acre by chainsaw and arranged some of the brush along the contour as a quick erosion control device, also digging a couple of swales to increase infiltration. The soil was well-drained, beautifully structured, and friable; I corrected for acidity and low nutrient content by scattering wood ashes over the area.  Then I began the slow process of figuring out what to do, and how to do it. 
 
It took me some time to figure out now how to lay out the site on the 25% slope; no bulldozing or any other kind of site leveling was going to be used here; I didn’t want to trash the topsoil or create water seepage problems that would have to be dealt with, nor to lose the heat gain generated by the southeast-facing slope. The site was a perfect “sun trap:” surrounded by woods on three sides and open to the south and southeast, with the hill to the northwest. It was typically fifteen degrees warmer than the chilly valley bottom even without a greenhouse. Lots of direct sun, very little wind—what more could a greenhouse grower ask for?
 
Once the site was laid out, I commenced cutting Norway maples for the side wall posts, to be spaced once every five feet along the total 95’ length. Why Norway maple? Because it was mandated for removal from the property under a forest stewardship plan devised by the Maryland Department of Natural Resources (DNR) in 1998, in exchange for which the community paid lower taxes than otherwise would have been the case. The Norway maple is a dreadful forest invasive which, left to its own devices, can out-compete natives by intense shading and secretion of strong growth-inhibiting substances, leading the creation of single-species stands of trees with absolutely no understory—and no leaf litter cover, either. I found an eroding hillside under a full stand of trees, nothing but bare ground—a shady green ecological desert. After consulting repeatedly with community members, I began carefully to thin out some trees, selecting for the straightest poles, leaving enough canopy so as not to enable the multiflora rose to go wild underneath. Such is the state of our younger forests! I then dragged the cut poles out of the woods and with the aid of Nick Corso, a young community member, and his pickup, we transported the posts to the construction site.
 
Later, we lit a huge bonfire in a conveniently located old pit (the remnants of an old sun-pit greenhouse) and “roasted” each post in the fire for a few minutes each, so as to protect them against rot and termites. We then dug post holes and set the posts in the holes, of course setting them as plum as possible given the fact that they were only as straight as unmilled lumber can be. Later I shot two levels, one for each set of posts on each side, notched the posts along the level with a chain saw, and affixed doubled 1x4 pine headers to connect the posts and prevent them from settling from one side or another.
 
So, there we were with two walls of posts. How to make a wooden roof than would span twenty-five feet across, be steep enough to shed snow load, sturdy enough to support whatever load it was likely to bear, including wind pressure, and not so heavy as to overload the side wall posts?
 
Fortunately, I was not without help, and so didn’t have to figure it out by myself. Heathcote includes several members with construction and/or engineering backgrounds—Chris Madden, who helped me get the site laid out and provided many caveats for the various designs which were presented; Mike Graff, a former engineer, who helped brainstorm and critique the various roof truss designs and helped supervise our various community “work days” when I needed multiple sets of hands for certain hands; and Bob Geissel, a retired building contractor with a brilliantly creative streak, who came up with the final roof truss design and provided many of the tools and much of the training which enabled my helpers and me to implement the design. 
 
Bob came up with an “I-beam” design which consisted of two pine wood flanges with ¼ inch plywood webbing in between. The pine was provided by Nick, who scavenged it from a shipping yard in Baltimore; the wood had been used for packing and came with all nails removed. Nick just went down and collected it in his pickup and a trailer, then stacked it and covered it. Thus we were able to stay true to the parameters of keeping the greenhouse construction very low cost and using local and/or reused materials.   We did pay Nick for his materials, mileage, and labor; still it was very cheap.
 
The I-beams proved to be wonderfully sturdy and lightweight. They were fastened at the peak with gussets and strengthened with collar ties about four to five feet down. The downhill side was about nineteen feet long, the upper side about twelve feet long, to cover the distance and still provide a steep enough angle to shed snow efficiently. The only downside to these trusses was that the I-beams proved exceedingly labor-intensive and time-consuming for us to make. Once made, however, Bob came up with an extremely easy and efficient way to mount them—place the ends loosely on top of the headers next to the wall posts, then push the peak up to a waiting roof person (myself) by means of long push-poles with forked sticks on the ends so as not to let the peaks of the trusses slide off the poles. Thin strips of pine were nailed on near the peaks of the trusses so I could nail the other ends of these strips to the preceding truss, thus providing a loose temporary connection between the trusses. Later we switched to using a rope by means of which I pulled the trusses up to me from the ground, with the ground crew setting them on the headers and helping get the trusses off the ground. It took a crew of about five people to accomplish this efficiently and safely. The ends of the beams were then fastened to the posts with long spikes with the bottoms sitting on the headers. 
 
In two sessions we thus set the twenty roof trusses in place. I then attached a permanent roof on top of the trusses, a few inches from the peaks, again using doubled 1x4x7 ½ pine.
 
To slow decomposition, no pine has been set on or near the ground and all pine has or will be treated with stain. The uncharred portions of the Norway maple posts have also been stained, with the cracks heavily caulked.
 
I selected some recently dead or moribund black locusts for use as end wall posts and with Nick’s help again we transported them to the site. We set these indestructible posts it the ground without any treatment. Incidentally, we have used no concrete, only a layer of rocks at the base of each hole to slow settling. The end trusses were bolted to the end posts and some pine cross pieces nailed onto the posts to be used as nailers for the plastic. Finally, on Bob’s advice, I ran a set of doubled sixteen-foot two-by-fours from each of the central end posts, a few feet up from the ground, to the peak of the third truss from each of the ends. These, together with the roof ties, prevent the trusses from being blown down like a set of dominoes. 
 
Last but not least, some of the other interns and I set to collecting some old carpets and cutting them into strips several inches wide, which we nailed on top of all upward- or outward-facing wooden components, especially the tops of the trusses. This was to prevent the plastic from being torn when we stretched it over the structure. I’m happy to report that it all worked perfectly. On January 21, six months after I began sawing the first Norway maples, we stretched the plastic over the structure and nailed it on using carpets and thin strips of ¼ inch rot-resistant Luan plywood as battens. 
 
One 13” snowfall and several windy days later, the greenhouse is standing tall. It still needs permanent doors, some more cross-bracing between the longer I-beams, and some more painting, not to mention terracing, irrigation, and installing roll-up devices for the side wall plastic.  But the thing stands, and it holds heat; it’s nice and hot in there on sunny 26-degree February days. Total cost so far comes to around $1600 and we are nearly done.  For  a 25’x95’ timber structure, that is truly minimal. The labor cost, however, has been huge, partly because we have been all figuring out how to do this as we go along. And to begin to make a living at this high tunnel growing, just to be able to rotate crops between them, I would need at least two more! At least one of them will be just downhill from the one we just built, on the same kind of slope.
 
For the next structure I am therefore contemplating a simpler roof truss, using light pole wood, along the lines of a primitive lean-to or pole barn. It will probably necessitate limiting the width of the structure to no more than twenty feet or so. Nonetheless, we are well on the way to getting a whole lot of produce out of an extremely limited piece of land. Wish us luck, and I hope this tale can inspire a few others to make the most out of what little they have to increase both economic and ecological sustainability in their lives, and to have a great time doing it—and to learn.
 
By Matthew Belmont
This project was supported by a $1500 grant from the Koinonia Foundation.