Sunday, 23 January 2011

The Hide

The modular sousterran idea is all very well but I have to survive somewhere until I have planning permission to build it. I can't afford to rent, so I'm pretty much going to have to live – while I'm at home, at any rate – on my land. This design exercise is to see what is the cheapest and least conspicuous living space I think I can cope with through two winters. Cheap obviously means small, but surviving through winter means reasonably well insulated.

This design is fundamentally based on my present bed, which is an IKEA loft bed with a desk and bookshelves underneath – a quite cosy and comfortable working space. I started from there and thought, 'OK, how much more do I actually need'. A design constraint is the spacing of trees in my wood. They've mostly been planted at pretty exact two metre intervals, although the rows don't precisely align. I can, of course, cut trees down – it's my wood, and, furthermore, it needs to be thinned – but in the interests of hiding the hide I don't want to cut too many down. It won't be good for the planners to know I'm living on site while I'm applying for permission to do so.

The plan is a hexagon of side 1200mm – basically, that's the biggest hexagon I can fit onto two sheets of plywood. I could get a bit more space by using a 2400mm cuboid, but cuboids are ugly and that's 25% more wall. Also, it's easier to fit the hexagon into the wood than the cuboid, which would definitely need trees cut down.

It's beneficial to have the cabin off the ground to avoid damp. Having it head-height off the ground means you can use the area under it as sheltered storage. There should be no need to kill the trees that support it, two poles lashed crosswise to four living trees should be fine.

The downside of the hexagonal plan is the bed ends up an odd shape – it is 2400mm long on its long side, but only 1040mm wide and only 1200mm long on its short side. It's big enough, I think, provided any concubine is friendly!

The hide as designed has a wood stove for heating and basic cooking, but there is no lavatory, bathroom or laundry – those facilities will be available in the Void. A stove is quite expensive, of course, but I think it's essential.


My idea is to construct the hide as two modules and six roof panels off site, under cover, transport those modules to site without their insulation and final outer cladding, erect them, insulate and clad them. The module structure will be primarily WBP plywood with softwood framing, assembled using the wood epoxy saturation technique, so pretty durable. Insulation will probably be 100mm fibreglass felt, although if I can find something more ecological that I can afford I will. All panels, including the floor, will be insulated.

The rear module will contain the built-in desk, bookshelves, cupboards and  bed. The front module will be considerably simpler, containing the bunk ladder and wall cupboards. The modules will bolt together – the reason for applying cladding on site is that the cladding will cover the bolts. Also, of course, the unclad modules will be lighter and easier to manhandle.

The cladding will be tongue and groove softwood weatherboard painted with an olive drab exterior preservative paint, but nevertheless is not expected to have the same life expectancy as the core structure – it can be replaced if needed. In order to prevent nailing the cladding on from breaching the encapsulation of the epoxy-protected core structure, sacrificial 25x50mm strips will be bonded on for the cladding to be nailed to.

It would be much simpler and cheaper to make a lower, hexagonal roof. Making the roof conical is an architectural conceit. The roof will almost certainly have to be cut on site to make room for the trunks of the supporting trees. It is built as six panels each having a flat sheet of plywood as its inner surface, a flat plywood soffit board, and an outer surface planked with tapered softwood planks. Between the inner and outer surface will be insulation as on all other panels. Some arrangement for ventilation will be made at the eaves. One of the panels will have an aperture designed for the stove pipe to pass through. After the hide has been erected the roof will be covered with tarred felt, and the final top cone, probably of stainless steel sheet, will be fitted.


The hide described in this paper really is small, and in cold wet weather is bound to be claustrophobic. Nevertheless with a small wood stove in a small and well insulated space it should be cosy, even in very bad weather, provided the door fits well! In better weather, and to entertain anyone, it needs to be extended, and this can most simply be done by providing an awning. An awning to provide a 5 metre square sheltered outdoor space would mean felling four trees.

It's probably not possible to provide a cludgie in the wood as the drainage ditches all connect back to the Standingstone burn and are sufficiently closely spaced I doubt you could get the requisite distance from running water. I'll check this, because some sort of loo would be a good thing.

A stove of the sort used in yurts would provide adequate basic cooking; in practice except for baking bread I don't often use an oven. For more elaborate cooking we will probably eventually have a cooker in the Void.

I'm not sure about electricity – some form of light is going to be necessary in winter, otherwise it's going to be a little grim; and it would be nice to be able to use my little laptop without having to go up to the Void every day to recharge it. But solar panels are not going to work in the wood, and mains is not on for all sorts of reasons. Even a small wind generator would attract attention.

There's obviously nowhere near enough storage space in this for all the clothes I use regularly, or all my tools, computers, books and things. This is minimalistic, temporary living, and most of my gear will have to be stored in the Void.

Thursday, 20 January 2011

Sousterrain revisited: there is a plan B

Having considered my earlier note on the design of a sousterrain for a month, I'm now going to rip it up and start again.

Reasons for not building in concrete

The first reason for not building in concrete is obvious. The embodied energy is huge. To present a largely-concrete sousterran as an energy efficient or 'green' building is hard to justify to me, to the planning authorities, or to anyone else. However, there are two other, pragmatically more compelling, reasons not to use concrete.

The first is that I have limited experience of using concrete, and that, once constructed, a concrete structure is very hard to modify.

The second is that to have any guarantees of the integrity of a concrete structure, the material has to be of a consistent mix and individual modules have to be poured and cured essentially in one operation – if part of a module is part-cured before another part is poured, you will have lines of weakness. That's not a problem if you use readimix. However, my proposed site is quite a long way from any road that a readymix truck could comfortably use, and I have no plans to construct such a road (nor can I afford to). And, even if I had such a road, it passes under electricity lines which do not in my opinion have safe clearance for something as high as a readymix truck.

Reasons for building in timber

Timber is in principle a sustainable resource. Its processing does not embody very high energy inputs. It locks up atmospheric carbon for the lifetime of the structure. As a building material, it's as green as it gets. Furthermore, it has other advantages. It is (relatively) light, strong, and easy to work. If you want to modify a wooden structure, it is generally easy to do so. Its disadvantage is a corollary of its advantages: it is bio-degradable. In the long run, it rots, losing structural integrity and releasing the captured carbon back to the atmosphere.

However, there is are ways of working around the rot problem. One with which I'm very familiar is epoxy encapsulation Done right, provided the encapsulation is not breeched and the epoxy is not exposed to sunlight, it prevents rot indefinitely. Of course epoxy is (at least at present) synthesised from fossil hydrocarbon. Furthermore it's at least conventional to use glass fibre as a reinforcing material in some parts of wood/epoxy structures, and in the case of the sousterran a layer of glass fibre cloth laid over the dome sections would help to prevent localised pressure points causing an encapsulation breech, while a tensile belt of glass fibre tape would prevent the dome spreading. However, the embodied energy represented by the epoxy and glass in a wood/epoxy composite structure is orders of magnitude less than that in concrete.

My original primary reason for not choosing timber is that if a structure is built underground and starts to rot, it seemed to me that it would be very hard to repair. But the truth is that this isn't going to be very far underground – a metre at most. In the event of problems, uncovering the problem area in order to repair it is not in fact a very big issue.

Lightweight structure

The original plan for the concrete structure was that it would be cast in components off site and then moved on-site for erection. In concrete this was somewhat implausible – individual components would weigh in tons, and would require specialist equipment to move. Boy Alex has such equipment, but even so it would require a substantial input of his time and equipment. But building in wood, it becomes much more practical. No component should weight more than at most 200Kg, meaning that it can be managed by a team of men and simple hoists. Building components off-site in the Void means that they can be built out of the weather, and consequently the encapsulated timber stays very dry.


The problem with building a lightweight timber sousterran is that actually you can't. If burying to an average depth of 1 metre, the overburden is about 1.7 tons per square metre, or 42 tons per dome. I'm finding it hard to believe that I can engineer a structure which can handle that amount of mass. So while a turf roof certainly is possible, and an overall finished shape that looks natural probably is possible.

Temporary structures

One of the driving issues behind this note is this: before I have planning permission to build my permanent structure, I need to have somewhere to live. Furthermore, I'm now going to have very limited money – far too little to enable me to build my whole structure in one phase. I need a warm and weatherproof structure before next winter. I have considered buying a second-hand mobile home, or buying (or making) a yurt, or building a temporary cabin. Each of these would make a very considerable dent in my available funds. Why spend money on an essentially disposable structure, when I could spend it on part of my permanent structure?

The alternative is to build one dome of my permanent structure, and to add to it as funds become available. The problem is that it's going to take some time to get planning permission, that for a structure as radical as this anything which upsets the planners is going to make permission even harder to get, and that pre-empting planning permission is one well-known way to get planners' backs up.
However, if the structure is modular, and built of a (relatively) light material, it does not have to be initially erected in its final place. It could (with permission from my co-conspirators) even be erected temporarily inside the high-slat shed. Furthermore, building one dome early allows it to be treated as a prototype. In the process of building it, other, alternative ways of doing it better are bound to be learned.

A further advantage of having a prototype dome is that there is something physical to show to the planners and building regulations people, who are bound to be at least sceptical about the structure. Also, if they do insist on changes, only one dome needs to be changed.

Hybrid structure

Both this note and the previous one have essentially presented the concrete structure and the wooden structure as alternatives. It isn't necessarily so. Some degree of concrete foundation is likely to be necessary – even if it is only piles under the bases of the pillars. A pile-and-slab concrete foundation is also plausible.

Casting the pillars in concrete is relatively unproblematic – certainly very much easier than the flying buttresses or the domes. Building the walls in concrete blockwork is trivial (although it's easier to route services inside a wooden wall than inside a concrete one). Even the lintels joining the pillars are relatively simple to cast in concrete, whether they are cast in place or cast off-site.

So it would be entirely plausible to have a structure which was concrete with wooden domes, or concrete with wooden lintels and domes.

The flying buttresses in the original design then become an issue. With wooden domes having a tensile band, they aren't strictly necessary – the tensile band makes the dome self-supporting. But, sculpturally, I like them. They could be built in epoxy encapsulated timber; or they could be solid oak (which would be nice – a little expensive, but not dreadfully). Or they could be concrete. It's not a decision which needs to be made in a hurry.

Note that the hybrid structure doesn't work well with the idea with the idea of erecting the prototype dome off-site, and, for that reason, doesn't work well with the idea of getting a prototype dome up quickly to live in in the short term. However, if the pillars and walls of the prototype had to be sacrificed that is not necessarily a huge loss.

Plan B

So, plan B is as follows.
  1. Start immediately (in March) to build one experimental dome in epoxy encapsulated timber with an intention to have it habitable by May
  2. Erect that in a suitable place 'off site' (and not earth sheltered) until planning permission has been obtained.
  3. When planning permission has been obtained, dig out the platform (the entire platform large enough for all four planned domes).
  4. Lay suitable foundations for one dome.
  5. Disassemble the prototype dome and re-erect it on-site.
  6. Back-fill over that dome only, leaving the remainder of the platform clear.
  7. Occupy that one dome, at least for winter 2011-2012; build other domes in a similar fashion at funds allow.

Tuesday, 18 January 2011

The joys of data transfer

OK, so, at this stage the main thing this blog is for is to find out how to import existing blog posts into Blogger. Brief summary: my existing blog uses a blogging engine I wrote myself back in 2000; it's quite a good blogging engine but it's not used by very many people because I didn't promote it enough back in the day, and so it's time to end-of-life it and migrate the existing users to something else.

Blogger has a mechanism for exporting and importing blogs. So, I thought, it ought to be possible to generate  the export format, which is documented here, from my existing data and then import that. From the documentation it was clear that the format was slightly bizarre - a well formed XML wrapper around data which is actually XML and presumably also well formed but is represented as text. However, I generated stuff that looked right to me according to the documentation, and it failed to import.

Worse, the error message given was terse to the point of unusability, and there's apparently no documentation of the error codes available on the web.

So the next thing to do was to export this blog from Blogger, and see whether the export file format looked anything like the documentation. And, guess what, it sort of doesn't. Which is to say that while the file format shown in the documentation is a very small subset of what's actually generated, it is such a noddy example that it doesn't even nearly represent what one needs to generate.

And, of course, there's no guarantee that even if I did succeeded in generating all the cruft that's in the format as generated by Blogger, I'd get my data to reimport - since the format includes magic identifiers which may represent objects in Google's persistent data space.

So it may be time to think of other ways to work around this.

Monday, 17 January 2011

End of eating dogfood

I wrote the Press Release System (PRES) back in 2000, and have kept my own blog on it since 2004. It wasn't a bad blogging engine. But I didn't promote it nearly enough, and very few people use it. So it no longer makes sense to maintain it - it's time to move to a more mainstream engine.

I can't help being slightly sad.

Creative Commons Licence
The fool on the hill by Simon Brooke is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License