First conceived of as a covered corridor along the back of the building with a bit of space for outdoor clothes/shoes and a small toilet, its primary purpose was to prevent rainwater falling on the ground immediately behind the building and running straight down the back walls of bedrock in – and into – the two lower rooms. With a roof to keep the ground dry, we would then be able to dig a drain into the bedrock itself behind the covered area to catch runoff from both the roof and slope (surface and subsurface) and carry it away from the back of the building before it reaches it.

By doing this, I wanted to avoid all that stuff with deep digging, tanking with damp proof membranes and concrete, etc, that tends to be the modern solution to water ingress in buildings built into slopes. Stopping the water reaching the back of the building in the first place seemed a much simpler solution. It keeps everything breathing and flowing rather than backing up behind a mass of concrete and plastic which, in any case, would be a mission to accomplish given that the walls are built straight onto bedrock which itself forms part of the structure, so no isolation of the building from its immediate environment is even possible.

And if, after our efforts, some water does work its way right down through the bedding planes of the schist and eventually finds its way into the rooms, then it would be so small an amount as to allow simple evaporation to take care of it. We might even be able to use it to our advantage – for instance, in the storeroom where evaporation from the rock surface would provide a cooling effect and a level of humidity conducive to storing certain root vegetables, so the plan is to build root vegetable storage boxes along the back wall, with shelving for things that prefer a dryer atmosphere kept to the opposite side of the room.

But having thought it all through and arrived at what felt like a reasonably watertight theory, it kind of kept growing …

When it came to construction materials, I started thinking of a living roof covering, and from that it seemed logical to extend the roof right back to meet the slope, so slope and roof would eventually become visually indistinguishable. And to put in 2 drains for good measure: a gully where the roof meets the slope, and one at floor level beneath to catch what might manage to find a way through.

Once work got underway, taking the slope back to bedrock in order to cut in the lower drain opened up even more space. So we’ve ended up with an area big enough for a lot more than originally planned … log storage for stoves in the two upper rooms, shower as well as toilet, generous area for outdoor gear, space for a methane digester to supply a small amount of gas for cooking when we only want to boil a kettle … the possibilities are many.

The schist floor being laid

First eucalyptus timber supports and main beams go up

The beauty of plain round pole timbers against stacked dry stone … though not for long: this wall will be insulated with woodchip light clay and lime plastered between the timbers to enclose the wall’s thermal mass in the interior of the building.

Mid February and the early morning sun highlights the concrete pillars cast to support the main beams where there isn’t bedrock for the purpose. Wayne starts building the rear schist wall between rock outcrops.

First set of rafters and purlins. The entire supporting structure is being constructed in eucalyptus cut from our own woods (with a few rafters kindly donated by the previous owner of this quinta who still owns some woodland next to ours).

The rear schist wall meets the wall plate

Walls curve to follow the line of the bedrock

The first planks go down

A temporary waterproof covering gives a sense of the enclosed space. About a third of the way along now.

João linseed oiling the timbers. All timbers were first treated with borax (disodium octaborate tetrahydrate, not sodium tetraborate decahydrate) and will have 2 coats of linseed oil.

One big shed!

Jonny laying planks

Last of the full length sections

Max, joiner’s apprentice

Nelly’s new look-out post

Planking complete

Huge thanks to Jonny, Wayne, Lester and João for great work and equally great craic. Building should always be such fun! The only trouble is, now we’ve created this beautiful space, it seems a shame to start filling it with walls and divisions …

But plenty of time to think about that yet. The next stage is to place the final upright to support the purlins, complete the drains and check their run, install the lighting system – PET water bottles set into upstands in the roof – lay the waterproof membranes, install roof drainage and containing barge boards and finally lay the substrate for the living roof.

Log store patio roof under construction – membrane goes down on screeded roof

Schist slabs being laid

Log store roof complete and low walls for seating under construction – there will eventually be a fire pit in the centre of the patio

This store will provide about 12 cubic metres of firewood storage space plus room for a workbench and tools. The wood stored here will be used primarily for the rocket stoves in the kitchen and bathhouse/greenhouse, so another storage area is needed for wood for stoves in the upstairs rooms. This dovetails nicely with the final stage of groundwork and construction needed to keep the main building dry and to provide space for a toilet, a connecting corridor between the upstairs rooms and hanging space for outdoor clothing and footwear.

So the next project is to construct a lean-to roof, along with natural drainage, along the back of the building. A lean-to roof with a difference. We plan to take it right out to meet the slope behind the building, varying the pitch as necessary, and cladding it (after waterproofing) with straw and a light covering of topsoil as a growing medium. It will be left to seed naturally with the pioneer plant species – mostly carqueja (Genista tridentata), wild lavender (Lavandula stoechas) and various heathers – that thrive in the thin dry soil on the slope above it, eventually resulting in a roof that’s almost indistinguishable from the surrounding hillside. So really neither a turf roof nor a green roof, but a living, growing roof nonetheless. Minus brambles and pine seedlings though …

Slope behind the house cleared of mato and brambles

View from the other side

Two days ago, the first day of January’s waning moon, we cut some poles from the few eucalyptus trees we have growing in the woods. Apparently if eucalyptus is cut during January’s waning moon, it has much less tendency to split and twist as it dries. Although most of the eucalyptus grown in Portugal goes to make toilet paper, it can make an incredibly strong building timber if it’s seasoned carefully and kept away from any source of wetness. It’s also naturally resistant to the indigenous wood-boring insect population, so this is what we’re using for the framework for the back roof. We’ll be using the poles in the round, as we have elsewhere in this project.

Eucalyptus poles, freshly cut

7m eucalyptus poles for the main beams, stripped of bark

To minimise the chances of the poles splitting at the ends – the end grain dries fastest – we’ll be painting the ends with diluted PVA to slow the drying process. (A handy tip thanks to Jonny from Quinta das Abelhas, the newest member of our now 4-strong part-time construction team.) As it happens, the ‘ecological’ wood preservative we used for the main roof timbers contains PVA, so is ideal for the purpose. The time of year is also on our side as we have another 3-4 months or so of slow drying in cooler temperatures and moister atmosphere before the summer heat hits us.

All the poles stripped

Wood for a roof, and all harvested from our own woods!

Burning the bark

Once the roof is completed, I’m intending to insulate the back wall of the house externally. The building gets no sun at this time of year so we have no possibility of utilising passive solar gain. In some ways, this actually makes designing for the extremes of temperature here much simpler. The rear walls contain by far the greatest proportion of total wall area in each upstairs room, roughly equal to all the other walls combined. External insulation allows their thermal mass, together with the dividing wall between the upstairs rooms, to be used for heat storage in winter, conserving the heat we generate from the woodstoves for as long as possible. This will significantly improve winter comfort levels and energy use in the building and, with internal insulation on the remaining outside walls, prevent walls at ambient temperature from sucking all the heat out of the room. Anyone who’s lived in an old stone house without adequate insulation will know this only too well …

In summer it works the opposite way round, keeping the building much cooler, especially with the benefit of the solid rock back walls and floors on the ground floor.

I’ve looked at various natural insulation materials and focused on ‘slip-chip‘, or wood-chip light clay. It’s a flexible material with good insulation properties and a wide range of possible ways of using it. We’ll need to experiment to find the best application method – either plastering it directly into and onto the stonework or pouring into reed mat forms fixed to roof timber supports – but the raw materials can be sourced very locally and very cheaply, if not for free. A final coating of lime plaster will provide a breathable finish, and a more durable and harder one than clay plaster (which will be used internally) in an area of heavy traffic where the walls are much more likely to be frequently knocked and scraped. A smooth white-painted wall will also enhance natural light levels in the corridor, which will be lit in daytime by 2 litre clear plastic soda/water bottles filled with water fitted into the roof.

Being out of sight under the turf roof, the insulated and plastered wall won’t detract from the external appearance of the building’s traditional schist stonework either.

Well … that’s the theory anyway. No doubt I’ll stumble on many things I haven’t thought about yet along the way …

Some of the trees on the slope above the track have been harvested recently. Inevitably this has destabilised the surrounding trees which no longer have the protection of the mature trees that were sheltering them before. Trees will always come down in storms, but the soil is so thin and impoverished by this exclusive concentration on nutrient-hungry Maritime pines (Pinus pinaster) and Eucalyptus (Eucalyptus globulus) and the lack of effective soil-reinforcing understorey, that the trees have very little to hold onto. Landowners are, in fact, encouraged to cut the understory as it’s seen as a fire hazard. But it’s the trees that are the fire hazard, being two of the most combustible species on the planet. The natural vegetation of the area – oak, holm oak, chestnut and hazel and understorey of the Strawberry tree or medronheiro (Arbutus unedo) and other shrubs – encourages development of a deep moisture-conserving litter layer which not only adds to and recycles nutrients back into the soil, but helps prevent the tinder-dry summer conditions which encourage the devastating forest fires this region is prone to.

Finished installing the flue today in between showers. Should be finally able to connect up the stove tomorrow. Why does everything always take about 10 times as long as you think it will ?! Short answer is if I will sleep in late and fail to get over to Benfeita until lunchtime, then there’s only so many hours of the day left. Perhaps it’s a side-effect of this back-to-nature lifestyle that the urge to hibernate over the darkest days of the years become almost irresistible … Well that’s my excuse anyway.