Monday 29 November 2010

Site specific low cost housing for a windy site


Part of the issue of building housing on the site we're considering is the wind speed, which is high. I imagine we're all going to want homes which don't take a lot of energy to heat, and the wind-chill effect on exposed walls is going to be considerable. The parts of the site which are most exposed to the wind are also the sunniest – the southern and western slopes. Of course, one can insulate, and straw bales are worth considering.
However the alternative, given that we have reasonably steep slopes, is to get down out of the wind. And if you do that you also lessen the landscape impact of the dwelling dramatically.
We all need structures which are low cost and simple to erect (since we're likely to be using mostly our own labour).
My suggested response to this set of problems is to dig into the hillside, build structure, and backfill over it. The modules I'm suggesting are hexagonal, simply because I like the shapes sculpturally – cross vaults might be more practical, because you'd get a rectangular floor plan, but you'd also end up with straight lines on the hillside which would be more visually intrusive. One advantage of the hexagonal module is that you can put the modules together to follow the contour in a more flexible way. The approximate size of the module is 2.4 metres per side, or 4.8 metres diameter. This gives a floor area of 15 square metres (about 150 square feet) per module.
Note that I'm guessing that this structure would be low cost – I haven't had it costed by anyone remotely qualified to do so.

Sustainable?

This proposal is mainly concrete. I would have preferred timber but am not confident I can design in timber in a way that would be durable and maintainable under ground. As such there's a substantial energy cost in making the cement. There's also embodied energy in vermiculite used as filler. Overall, though, it isn't hugely much concrete – probably not as much per living unit as in a conventional block of flats. Because there's no wind exposure, and all the exposed sides are oriented towards the sun, it should have adequate passive heating even in winter; any supplementary space heating needed, water heating and cooking can be provided by a stove burning fuelwood grown on site. Large opening windows allow for cooling in summer.

Visual impact

One of the main objectives of this design is to minimise visual impact on the landscape. Because the terrace retaining wall is half the height of exposed sides of the structure, and because the slope above the structure matches the slope below the structure, from upslope or downslope there will be very little to see until you're very close.
From further away there will be a horizontal visible feature in the slope, but because all the curves are organic it shouldn't look too visually intrusive or artificial.

Elements

The building is constructed of four different concrete elements, as follows:

Dome segment

Obviously, six dome segments are needed for each dome. These will be the heaviest elements, and most difficult to move; it would be preferable to cast them in place, by erecting the pillars first and then erecting the former where the dome segment is to be located. Even if the segments are cast in place, however, they probably need to be separate segments rather than a single monolithic dome because of expansion issues. The cross section of the dome is probably a catenary, rather than an arc.

Pillar

One pillar at each corner of each dome, obviously. Hollow pillars – even if they do not contain drains – will be lighter and easier to install without being substantially weaker, so the core of the pillar element might be a spun concrete pipe section. The pillars need only be two metres tall because the natural shape of the edge of the dome will in any case form a slight arch.

Flying Buttress

Domes have thrust: they want to fall down, and in doing so they push out radially at their base. This thrust has to be resolved. In the unexposed edges of the dome (which, when building into the hillside, is most of them) that thrust is resolved simply by backfilling the soil. On the exposed edges – which we want, because we want light an air into the dwelling – the nature of hexagonal cells is that some corners are convex, and some are concave. How many of which you have depends on the shape of the site. On each convex corner you need something to take the thrust of the dome, and my choice it to have a flying buttress. This would span a terrace in front of the building and transfer load to its retaining wall.

Eve

The eve element is essentially just cosmetic; it joins adjacent flying buttress or eve elements into a continuous organic curve, again helping to reduce straight lines in the landscape. The eve units are designed so that although they show a fair curve on the outside they have a horizontal on the inside, so that ordinary standard patio door double glazed units can be used in all the exposed sides.

Building method

Again, this needs to be run past people with more experience than I have. Preventing damp ingress into the dwelling needs to be thought about carefully.

Dig back into the hillside

Strip off the turf and retain it. With a mechanical digger, dig back into the hillside; obviously, choose a site with contours which minimise the amount of digging that needs to be done. Reserve the soil removed, we'll need it later. Reserve topsoil separately from subsoil. The rock on site is part of the Kirkcudbright Shales and is probably soft enough to break up with a digger, is actual rock needs to be removed.

Lay drains

Lay drains under the site from uphill to downhill; these will later be joined to lateral collectors on the uphill side to reduce water pressure uphill. Rainwater drains can simply discharge downslope – they're only carrying water which would naturally drain down the slope anyway. A separate soil drain needs to lead to some suitable treatment.

Level platform

Level roughly with hardcore; finish with sand.

Pour slab

Taking one hexagonal module at a time, lay a waterproof membrane on the sand with at least 2.5 metres overlap on the uphill side, shutter the hexagon, and pour a self-levelling concrete. I need advice on the thickness of the slab. It almost certainly needs to be thicker under the pillars.

Erect pillars

It's possible that the pillars should be hollow with a drain down the centre of each one, linked into the under-slab drains, to move water from over the structure to under it. However, there's an issue about how to achieve a watertight seal at the top of these, so unless someone else can solve it that isn't part of the plan. In any case Pete suggests that the pillars should be founded on concrete piles going down to bedrock.

Erect uphill walls

Between each pair of pillars on sides which will be backfilled, erect a breeze block wall.

Erect terrace retaining wall

In front of the structure along edges that are to be exposed, erect a dry-stane dyke about a metre high and about two to three metres away from the structure, with footings in the right places for the flying buttresses. The exact width of the terrace at any point should be chosen so that when the structure is complete and backfilled, the slope above the building should continue as a fair curve in the slope below the building.

Part-backfill the uphill side

Bring the excess waterproof membrane laid down when pouring the slab up the sides of the newly erected uphill walls. Backfill, installing lateral drains as you go. At this stage, backfill to about 50cm below the tops of the walls.

Erect flying buttresses, eves and (possibly) lintel arches

In order to resolve the thrust on the domes, the flying buttresses and eves must be in place before the domes are cast – otherwise they'll just fall down as soon as the armature is removed. However, not all the pillars are supported against thrust in all the necessary directions until all the domes are up – to some extent, once completed the domes will support one another. One solution to this would be to cast lintels between each pair of adjacent pillars. If this solution is chosen, they must be installed at this stage. Alternative solutions are
  • Embed a tensile steel reinforcing belt as low as possible in each dome;
  • Construct separate armatures for each dome and leave them in place until all domes are cast and cured.

Erect domes

For each dome that is erected, a wooden armature needs to be erected to support the segments until all are in place. This armature needs to be supported on wedges so that when the dome is complete, the wedges can be driven out and the armature disassembled and moved to the next dome. Before the armature is removed, the flying buttress and eve elements which support the exposed edges of the structure must be installed.
Because of the steepness of the lower slopes of the dome, some though is needed on how to prevent the concrete from slumping before it is cured. Some upper surface shuttering may be required, and thought needs to be given to how to secure this. Also, it will probably be necessary to erect some sort of gantry or scaffold in order to apply and smooth concrete over the centres of the dome – because certainly that will be out of reach from the edges while the concrete is not set.

Fit chimney outer

Cut hole(s) for the chimney(s) in the appropriate place(s) in the domes (or, preferably, leave them when casting). Install vertical concrete pipe over the hole to the final ground level.

Fill valleys between the domes

On a settled, dry day, or better still on the first day of a run of settled dry days, fill the valleys between the domes with a lightweight inorganic filler material such as vermiculite. The object of this exercise is simply so that from every point on the structure, when the upper membrane is in place water will naturally drain to one edge and will not pool; the valleys do not have to be completely filled by any means. This doesn't need to be done if a watertight means of joining the upper membrane to drains running down internal pillars can be found.

Lay upper membrane

Lay a waterproof membrane over the whole structure, overlapping the uphill sides by half a metre. Fold this overlap down over the overlap of the lower membrane. Sealing around the chimney and any internal pillar drains is an issue I haven't solved, but I imagine suitable mastics may be available.

Backfill

Backfill over the whole structure, first with subsoil, then with topsoil. Contour the backfill to form a natural slope. Turf over. Similarly backfill against the retaining wall on the downslope side.

Install glazing

Ordinary standard double glazed 'patio door' units are installed into the exposed sides.

Fit out interior

To taste.

Rough costings

The following costings are preliminary, and are based on concrete costs only, with no allowance for levelling the platform, fitting drains, services, labour and so on; it's mainly to cost the bits of the structure I don't have a good feel for. This looks startlingly low – as if I've got something wrong by a factor of ten. I hope not!

Sousterran: quantities and mass











Concrete option





Cladding




Constants
Unit



Item

Unit Price per sq m
Mass of 1 cubic metre of concrete 2400 Kg



Bituthene 3000

20 metre roll £163 £9.03
Mass of 1 cubic metre of soil 1700 Kg



Styrodur 3035CS, 30mm

14 sheets at 1250x600mm £70 £6.67
Price of 1 cubic metre of concrete 100 Pounds sterling



Styrodur 3035CS, 60mm

7 sheets at 1250x600mm £70 £13.33
Price of 1 8x4 sheet 15mm exterior ply 40 Pounds sterling



Total, with 30mm



£15.69
Price of 1 sq metre concrete blockwork 10 Pounds sterling



Total, with 60mm



£22.36
Dome radius (long axis) 2.4 metres









Area of dome floor 14.98 square metres









Area of dome surface 36.19 square metres









Number of wall panels 11 at 4.8 sq metres




















Elements Area sq m Thickness m Volume cu m Mass tons Concrete Cost Cladding cost Number of Total cost Total mass Shuttering sheets Shuttering cost
Floor 104.83 0.1 10.48 25.16 £1,048 £399 1 £1,448 25.16 1 £40
Dome 36.19 0.1 3.62 8.69 £362 £809 4 £4,685 34.74 42 £1,680
Dome segment 6.03 0.1 0.6 1.45 £60
24
34.74 7

Pillar

0.5 1.2 £50
17 £850 20.4 6 £240
Eve

1 2.4 £100
5 £500 12 4 £160
Flying buttress

2 4.8 £200
4 £800 19.2 6 £240
Wall 4.8 0.1 0.48 1.15 £48 £75 11 £1,357 12.67 0

Overburden 104.83 0.2 20.97 35.64 £0
1 £0 35.64 0 £0













Subtotals





50 £9,639 194.56
£2,360
Total, cost









£11,999


No comments:

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