gemini://ur.gs/blog/garden-room/

In my last post, I talked about a straw bale course I went on, and a few projects I had in mind coming out of it. One of them was the idea of a "garden room".

My wife very much liked this idea, and so did I - so we've decided to make it the first thing we tackle, and figured we could get stuck in as soon as we found a source of bales. She has a few friends who are farmers, one of them knows someone who makes bales, and, well...

I got one hundred bales. It works out at about 16 cubic metres of straw (top tip - 1M³ of straw more or less equates to one linear metre of wall). not all of it fitted in the garage - I've got an overflow pile in the garden, and we had to sort and throw away / gift / recycle a whole bunch of stuff that *was* in the garage. If I can't use it, things might get a bit strained at home ^^.

The bales are barley straw, very similar to what we worked with at CAT; as there, most of the bales are ~900mm long, with a few much-longer outliers. I don't have a suitable moisture meter, but they feel dry - the farmer's son said they monitor moisture when baling, and stop if it goes above 15%, and then they're stored up high and under cover, so I'm not worried by that. The strings are nice and tight, and their density is ~105kg/M³. That could be higher, but it's certainly good enough. Best of all - no wasps.

If you need a source of bales in Yorkshire, do get in touch and I can share their details. I'm afraid it'd be a bit too much like posting my own home address to name them publicly. Farmer and son were both very interested in the building process and want to be kept up to date with it, as do several passers-by who stopped to gawp as a trailer full of straw bales got unloaded outside a little house in suburbia. Definitely a conversation-starter, if nothing else.

I knew from the course that I didn't just need the bales. I also need plenty of wood, to build the baseplate and roofplate; material for foundations; roof material; lime plaster; clay plaster; etc. To work out what I need to buy, I needed a design. There's also the small matter of, uh, regulatory compliance. In England, a lot of building projects need approval from both Planning and Building Regulations. If they don't get it, an apologetic-yet-burly instrument of state force will come round and knock it down again, eventually.

Fortunately, there are a series of exemptions to both planning permission (called "permitted development", or PD) and building regulations. In my case, as long as I keep the building's height (measured from the highest adjacent piece of ground to the highest point on the building) below 2.5M, ensure the floor area is under 15M², and avoid things like beds, electricity, running water, etc, then I can claim permitted development under class E that is exempt from building regulations under class 6(3).

Planning and building regs have different definitions for things like height and floor area, but that's what the intersection of the two permits in my case. Your case may well differ.

This is actually quite limiting, more on the height than the floor area, but I've run into both constraints. You want at least 2M to stand up in, straw bales want to be lifted up at least 300mm off the external floor level to avoid getting wet, and you need to leave space for a roof - ideally with some sort of pitch on it. So the height gets used up quickly, both from the outside and the inside. But if you want bigger, you have to ask for permission, or resite the building, or perhaps both - not really options in my case. Time to design within the constraints.

CAD is very much not my forté, but there's a package called SweetHome3D that I can just about use. It's mostly focused on interior décor, but can be used to lay out walls and get an idea of dimensions, etc. Good enough to get a rough idea, at least. So I had a quick measure of the available space in the garden and started by knocking up this:

Thankfully, I revisited this a few times before making any decisions. In particular, it's very important to explicitly model the *thickness* of the bales. For a small room like this, on a constrained plot, they eat up a surprising amount of the total area. If I had a particular minimum size in mind, I could see that being quite irritating, but this is just going to be somewhere to sit and watch the outside, so almost any size could work.

The garden room is replacing an area of mostly-rotten decking with a pergola over it, supporting a huge viney plant that's sort of nice, sort of annoying. The brick wall I'm building against has a northeast orientation and the back of the pergola also butts up against it; it forms a square around 2.4x2.4M. There's an apple tree at the NW corner, and a much-more-functional, independent piece of decking at the NE corner. Between the two, there's about 4.6M of usable space (so the pergola covers about half of it).

After taking the straw bale thickness into account, a 4.6x2.4M (exterior) garden room with a curve in it turns out to have very, very little internal space. I reworked the D shape into a much gentler curve and wound up with around 5sqm internally. So, to do anything large, the vines have to go. I was kind of hoping the pergola footings would be useful, but they're in the wrong spot. Pushing to about 4.6x3M gets a much more reasonable footprint. Sprinkle in a few doors and windows and, ta-daaa:

Sweet Home 3D rather conveniently ignores foundations and roof, but of course, I need them before I can do anything else. Straw Works do a bunch of "standard details" that are "tried and tested", and Bee's motto is "don't reinvent the wheel" - we're steered to re-use these as much as possible.

A DIY project like this would normally use something like the "rammed car tyres with suspended timber floor" detail, which avoids a lots of effort in the ground. It has a "structural box beam of relatively heavyweight timber, which I'm certain is overkill for a project of this size, but as a starting point, it's a lot easier than the plinth wall details - and they require a higher level of skill.

The problem is, I ruled out car tyres - I don't really want them leaching stuff all over the garden. So I'm left with doing the plinth wall or coming up with... something else. This has involved learning some physics.

The mass of the building rests on the baseplate, which rests on the foundation, which rests on the soil - which eventually sits on bedrock. Your first question is how much the building is going to weigh - and straw is not super-lightweight. 105kg per metre of wall, about 17 metres of that, so about 1.8 tonnes. Add render to the outside and plaster to the inside, a lot of wood, the roof, and (for a suspended floor) the floor and people - not to mention the foundations themselves - and I'm probably pushing 4 tonnes or so.

How much force is all that? Can the soil support it, or am I going to need to excavate to bedrock? Fortunately, this one is easy to answer - once you know it. Force (in newtons) = mass (kg) * acceleration (m/s).

For a building just sat doing nothing, the acceleration is the force of gravity pulling it to the centre of the earth - that's what the soil has to oppose. 4000kg * 9.8M/s = 392kN.

The "bearing capacity" of different kinds of soil is publicly available. My whole garden is made ground, and quite variable, but digging below the topsoil there's a lot of clay and it's very hard, with stones mixed in, so I can happily go with 300-600kN/M². Since it's force per unit area, this is a measurement of pressure - like psi, atm, pascals, etc.

The units give a clue as to what's next - we have to work out the area we're spreading the building's mass over - the pressure it's exerting on the soil - which comes down to foundation design. The plinth wall option runs the whole length of the building walls (~17M²) and is 450mm at the base, so its area is 7.65M² - giving us 51kN/M². Even though the building is heavy in absolute terms, the soil can more than take it.

How about something plinth-y? If I imagine 17 plinths (keeping the gap between each at 1M), each with a 500x500mm cross-section, that gives me 92kN/M² - the soil I have is more than capable of managing that, although some soft clays might struggle.

Car tyres are plinth-y. I could also get plinths using blocks of stone, or pillars of brick or block, or cast concrete. Lots of options, all quite expensive (either environmentally with the use of concrete, or just in terms of cash) - the plinth wall would be even more expensive.

To cut a long set of developments short, I have four test piers up in the garden right now. I cut down to subsoil, placed "bags for life" - these are woven polypropylene at around 120gsm into the holes, and filled them with either pea gravel or an attempt at lime-stabilised soil (10:1 soil:hydrated lime - so around 20:1 soil:lime).

I'm not convinced my stabilised soil will set - i did a jar test and the stuff I used was still fairly rich in organic matter, and might also be too silty. The pea gravel seems fine, though - and is very easy to level. However, for those piers, the bag is structural, and that worries me a bit. If the bag splits, the pea gravel spills everywhere and the pier is gone.

What are the chances of that? There are two major considerations - forces acting on the bag, and deterioration over time. Car tyres are super-hardy, of course, and also proven by decades of use in construction. For my weedy bags, the first turns out to be easy to calculate, although it took me some puzzling at first.

If you take a cylinder full of air or water and squeeze down on it, the force is transmitted, fairly equally, to the whole surface of the cylinder. If you have a cylinder of a solid and squeeze down on it, that force is transmitted straight down into the base instead - the walls don't take any of it. A bag of pea gravel turns out to be somewhere in the middle of those two extremes, and the whole thing was modelled over a century ago by someone who wanted their grain silos not to fail. "Janssen's equation" allows you to figure this out for any given material at any given depth of the cylinder, but the maths quickly got harder than I could follow easily, so I decided to just take the liquid as a worst case and calculate based on that. Call it a safety factor.

Calculating the pressure on a single pier isn't too bad - 392kN divided by 17 pillars gives us ~23kN/pillar, and the bag surface area happens to be ~1M, so we can say that's the pressure too. So what can the bag take?

No idea, honestly. I cheated and converted the pressure number into something I had a better intuition for - atmospheres. It's around 0.atm of additional pressure inside the bag (so 1.2atm inside, 1atm outside). I can think about that and reason that if I blew air into the bag very hard, it wouldn't pop at that. It's around 3psi, 0.2bar, 20KPa. Another comparison - you can pressurise a 2L plastic bottle to around 6atm before it blows up. If anyone knows how to convert this intuition into a calculation, I'm all ears.

Longevity, I'm umming and aa-ing over rather more. 120gsm WPP sandbags are a thing, and apparently last a couple of years out in the elements, which is not super-encouraging. In the case of lime-stabilised soil, we're treating the bags with a highly alkaline substance. Pea gravel is rounded but not perfectly smooth - it has some sharp corners. I'm stabbing holes in the bottom of the bags to aid water flow. On the other hand, I have bags that have been at the allotment all year without any sign of degradation at all - and I already have lots of these bags. I can double-bag, etc. Again, if you know how to model this, I'd love to hear it.

Much better to over-design than under-design for foundations, so while I'm reasonably pleased with these piers, I'm going to try some "flexible bucket" pea-gravel filled foundations, and anything else I can find that's reasonably hardwearing. I've looked at animal feed troughs, water pipes, water tanks, bulk bags... I can't get the next stage going until I have foundations in place, so there is a bit of a rush on - I've already ordered the wood for that.

On top of a regular plinth wall, the StrawWorks details linked above just have a simple baseplate of 4x2 C16 timber. This isn't really intended to be structural, it just forms a nice base onto which you can place the bales, add fixings, etc. In the car tyre foundation detail, it includes a big fat "structural box beam" which is intended to bear the weight of the building over the gaps between piers. It's a lot of timber, and designed for a two-storey house with suspended floors and many appliances and people.

I popped the details of the 2x4 timber into a random timber strength calculator with the 1M gap between piers and it seemed perfectly happy that it could hold the weight with no breakage or deflection across the span - with loads of margin to spare, too. So I'm just going to do the plinth wall baseplate on top of the piers and hope for the best. If it starts to look sketchy I can always lift it up and put a proper box beam underneath it, of course. This is on order - I've got 10x4.8M lengths coming to the house as soon as they can arrange delivery, so I'll be able to do the whole thing without any joins mid-span.

If we use the 2x4 for the baseplate, then we have to use super-mini joist hangers and joists, since we only have 2" to go off, so this will probably necessitate pillars under the floor as well as around the walls. This takes some load off the exterior plinths, at least. I can't fit this until after the walls are up, since I'll need access to below the baseplate in all sorts of funky ways to do compression of the walls, etc.

As I noted above, I've already got the bales and they're nice. I've also sourced hazel and sycamore fascine stakes from Leeds Coppice Workers, which is a very nice-seeming co-op. £110 for 100, ranging from 1-2" diameter, and from 0.8 - 1.3M. All sharpened at one point as well, which saves me a bit of work. I do need to chop ~10 of them down into 40x 350mm stubs to poke up out of the baseplate and impale the first course of bales, but I can get onto that anytime. The shorter remaining stakes can be used at roofplate level, while the longest ones can be used at the four-bale level.

I haven't bought them yet, but trucker's straps are easy to come by. Packing straps are more difficult, at least at 4 tonnes of loading (this is the force of compression on the walls, which the packing straps are expected to hold long-term). I definitely don't want one of those snapping in my face.

There's some annoying detail to work out around the brick wall - talking to the course instructors and their friendly structural engineer, I might need to leave a ventilation gap and have Tyvek membrane over the straw. It might also be worth "french dipping" that face of each straw bale in lime slip and letting it dry before placing it, to give some weather and fire resistance.

On top of the wall goes the roof plate, which is built similarly to the baseplate. I'll order the stuff I need for that once I'm that far up - 2x4 is readily available.

I've got a pair of doors from the old shed I knocked down which will be re-used for this project. I'd also like some circular windows - perhaps up to three, although that might be one too many. I know how to put together the framing for them, but I'll need to measure carefully and, for the windows, order carefully.

Externally, we want lime render, which we won't be able to apply over the winter - so it'll be tarps or membranes until spring. Internally, a nice clay plaster.

I'm just not designing this yet. Too much going on in my head already. I figure I get the walls up, build lots of redundancy into the foundations and roofplate, then put whatever I can get away with up. Cedar shingle on a reciprocating roof would be amazing, but I expect not to have the space, and I also expect not to be able to get the necessary pitch (5/12). So maybe it ends up being an EPDM membrane with some kind of timber to hide it, or asphalt shingles (they're fine with a 2/12 pitch). If I really wanted to push the boat out, I could consider a metal roof - they can do super-low, almost-completely-flat pitches, but the materials are really not cheap.

The flat bale wall is protected by the brick of course, and we'll just need guttering along there , but the rest of the walls will benefit from the roof having a fair bit of overhang. Fortunately, nobody will ever see both front and back at the same time.

So that's where I am today. I've got a friend coming Thursday to help move the groundworks on a bit, and perhaps build some baseplate if the wood arrives in time. I've also got another offer of help towards the end of October if we're still working on it at that point. The weather's been poor the last few days, and if this keeps up we might struggle to make progress generally, but I'm hopeful we can at least get the foundations and walls done (perhaps not as far as compression, but I'd like to think so) and protected before winter gets here properly. If we can get that far, I free up a 7x5M tarp that can be a pretend roof for the duration; surplus Tyvek can protect the walls. Let's see how we go.

Garden Room

Bales

Design and materials

Foundations

Baseplate and flooring

Walls

Roof