The K-Zone: Kev's Kabin
This project got underway when I got so fed up with seeing the ugly spiked
metal railway fence at the back of our garden (our garden borders a railway
line) that I had to Do Something About It.
Since I had been toying with the idea of building an
office/summerhouse thingie for some time, it seemed reasonable to put the
building right at the back of the garden, and make it tall enough to obscure the
metal fence, for the whole width of the garden (about 17'). This
article describes the construction.
I'm not writing this to explain the best way to build a large timber garden
building --
after all, I'd never built one before either. Rather, I'm hoping that
other people will be able to learn from my mistakes, which were
prolific. It's also worth bearing in mind that my design accomodates
some odd features of our house and garden. In particular, the ground
has a pronounced slope -- about 6'' vertically for every 6' in length.
This is because our garden is, in effect, on the edge of a railway
embankment. There's an almost unlimited amount of soil and gravel
above the back of the garden waiting to slide down if I try to level
it out, so I've generally felt that it's better to live with the
slope than to try to fix it.
The other limiting factor that I have, and most people do not, is
a very restricted access to the garden from the road.
All building materials have to be carried through the house, negotiating
a narrow passageway between two rooms. So there would be
no prospect of bringing in a cement mixer, for example, let alone
a prefabricated building. The only way to
have a substantial garden building was to build it myself from timbers
and sheet materials.
Why build a garden office?
It is very unlikely that building a garden shed yourself, from
the kinds of
materials you can buy from a timber merchant, will be cheaper than buying
a standard prefabricated shed from a specialist supplier, even if you don't
cost your own labour, which will be considerable. Although many
commercial sheds are flimsy, you can get sturdy ones if you're prepared
to pay a premium.
Although you won't save much money, if any,
by building a shed
yourself, building a garden office is a rather different
matter. Commercial garden offices are designed for all-year-round
occupancy, which means that they need a high standard of weatherproofing.
It is necessary to include double-skin walls with insulation, pay close
attention to waterproofing and damp penetration, and install double-glazed
windows. Because the office may house expensive equipment (computers,
etc.) security is an issue. In addition, unlike a shed, a garden office
needs to have an inviting interior -- no rough-sawn timbers or
chipboard floors.
This is why a 16'x8' prefabricated shed or summerhouse can be
had for £1500-£2000, but for a prefabricated garden
office of the same size, prices start at about £10000 and go upwards
from there.
Now, weatherproofing, insulation, and security are partly a matter of
choice of constructional materials, but they are also related to
design, and constructional standard, both of which are areas
where a bespoke construction can score over a prefabricated building.
In short, one of my
design goals was to build a garden office for not much more than the
cost of a prefabricated shed of similar constructional
specification -- about £2000. This turned
out to be almost possible, giving a considerable saving over a commercial
garden office, even costing my time.
Regardless of the cost saving, if any, an important advantage of
building a large shed or garden office yourself, apart from the fact
that it gives you healthy exercise outdoors, is that you can have
exactly the design you want. In my case, I knew that the building would
be bordered by garden fences on three sides, so there was no purpose in
having windows or doors in those sides. I wanted plenty of door and opening
window area for good summer ventilation, but insulation and double
glazing for winter use. I wanted the door to line up with the garden
path, so I could get into the building without first walking across
the wet squelchy lawn. I also wanted something that looked a bit
different from the usual prefabricated shed, and I got that. Whether
it looks better or worse is, of course, a matter of opinion, but
it's certainly different.
Design and planning
My first design criterion was that the garden office would occupy
most or all of the whole width of the garden, and be as large as I could
have it without involving the local Planning or Building Control
departments. Under current UK
legislation, planning approval is generally not required for a
garden building that is less than 3m tall at the highest point,
and more than 5m from a dwelling and 20m from a highway.
But I discussed it
with the neighbours before starting work because, apart from the planning
issue, I didn't want to upset them.
I decided to make the base 16'x8' (or, more precisely 2440mmx1220mm),
with a 16'x4' deck to the front,
so I could use standard-sized materials with the minimum of cutting.
Cutting is straightforward with the appropriate tools, of course,
but measuring for cutting is time consuming. I could have made the
front-to-back length greater than 8', but the next standard size up
would have been 12', and that's just too big for my small garden.
In any case, under UK law an outbuilding must comply with certain
Building Regulations if it occupies an area greater than 15 m2
and is within 1m of the boundary. Presumably this is to reduce the risk of
fire spreading between properties.
Since my building is clearly within 1m
of the boundary, I wouldn't have wanted to make it so big as to require
the additional expense and hassle of Building Regulations compliance.
With the roof overhang, the area of my building is 14.4 m2.
My next criterion was that the roof should have a single slope from front
to back. This was for simplicity of construction, as well as alleviating
the necessity to deal with drainage of rainwater from more than one
end of the building. The roof had to overhang by at least 18'' at the
front so that the doors and windows could be opened when it was raining.
I wanted to make the roof pitch at least 1:4 so I
could use a single layer of roofing felt, and this turned out to
be just about possible,
while still having the
roof joists high enough to clear my head (6'4'')
at the low end.
The highest I could have the front of the roof was 3m above ground (planning
restriction), and the lowest I could have the back was 2.3m (my height
plus the thickness of the joist and roofing material, plus a nominal floor height of 300mm), which -- allowing
for an 18'' overhang at the front -- gives
a roof pitch of 1:4 exactly.
I did toy with the idea of having a curved roof (like the
garden offices made by Henley) because I thought it
would look nicer, but I suspected that this would be
difficult to achieve with timber (Henley's products have prefabricated
plastic roofs). In any case, it has turned out to be impossible
to see the roof (except from below) from any part of the house or garden.
Foundation
|
|
The foundation, skids, and bearers
|
There's no doubt in my mind that a building of this projected size and
weight ought
to have a poured concrete foundation. That is, I should have dug out a trench
about a foot deep, and filled it with concrete. Unfortunately, this was a
course of action not open to me for two reasons. First, the slope of the ground
is such that that, if the poured foundation were perfectly flat, its surface
would have to be about 18 inches below soil level at some points. Not only
would this leave me with the problem of disposing of several extra tons of
surplus soil, but it would put the bottom of the building well below
ground level.
This would make it even harder to keep dry than it would otherwise be. Of
course, if I had a large garden I could have levelled a much larger area than
the building's footprint, and dug a trench within that. Unfortunately, my garden is
small, and the building comes to within a few inches of three of my boundaries, so
this wasn't possible. Another possibility, I guess, would have been to have the
foundation below ground level as described, and raised the whole building on wooden
piers. This is the approach I adopted for the kiddie playhouse elsewhere in the
garden, but that's only 5'x5' so the amount of construction wasn't so daunting.
The other problem is that as there is limited access to the garden, and none
from the road, I would have had to mix several tons of concrete by hand. Now, I
don't mind hard work, but that's just too much.
So in the end I built the foundation in the form of 15 supports, in
three rows of five. Each support conists of a hole about 2'6'' across
filled with compacted hardcore and
gravel, with between one and four concrete blocks standing on it.
The number of blocks reflects the height of the land -- one block at
the back left of the site where the ground is highest, four at the
front right where it is lowest. As can (sort of) be seen from the photo, I've
concreted all around the piles of blocks on each support, so that each
support is, in effect, a single lump of concrete 2-3 feet across and 2-8 inches high.
Cost of this part of project:
Concrete blocks £100
Approx. one ton of sand, cement, and gravel £100
----
£200
Base, floor, and deck
The concrete supports are spanned by 8''x2'' skids, on top of which
are 4''x2'' bearers at 12'' spacing. The bearers are screwed to the
skids using steel L-brackets, and at the front another run of
4''x2'' timber is screwed to the bearers using 6'' coach screws.
I adopted this arrangement at the front of the base because the weight on the
front edge of the base is going to be enormous --
it will support most
of the weight of the overhanging roof, and the solid wood doors and
windows, and the substantial front cladding.
So my hope is that this arrangement will spread the load
across the supports more evenly.
|
|
The main part of the base built and treated with wood preserver
|
The building floor is made from four 8'x4' 19mm plywood panels. Note that the
panels span the middle of the base, because the front load-spreading timbers
can not (they aren't long enough). There's an expansion gap of about 3mm
between the panels.
I suggest that it's necessary to be quite careful about the selection and
preparation of the timber for the base. Not only will it be in a relatively
damp location, but once the structure is built it will be inaccessible for
repairs. I think it is important to ensure that no part of the timber is
in direct contact with soil, and that you've done everything possible to
treat the timber against rot, woodworm, etc. All the base timbers in
my building are pressure-treated, and have had at least two
subsequent coats of thick, waxy wood preserver. The problem with wax-based
wood preservers is that they take weeks to dry in damp, cold weather, but
I think it's worth the inconvenience. All the timbers were stood overnight
with their ends in buckets of wood preserver, as I think it is the grain
ends that are most susceptible to the entry of moisture. There's a vapour
membrane between the base and the ground, to limit condensation of groundwater
on the bottoms of the timbers. However, I think it's important to keep the
vapour barrier as far as possible off the timbers themeslves, as they
need ventilation to remove any condensation that does form.
|
|
The base with front deck under construction
|
Using top-quality exterior grade timber, and the 40 litres or so of
expensive wood preserver it took to treat them, probably added about
£150 to the total cost of the job, compared to using ordinary
sawn softwood and a couple of coats of cheap creosote substitute.
The deck at the front of the building serves as the step up to the door (which
would otherwise be a 12'' step, the ground slope being what it is), and hides
the ugly concrete foundation. Because it doesn't support a huge static load, it
is supported using the normal technique for garden decking, which is to bolt
the deck bearers to upright posts set in holes full of concrete. The black
cloth that can be seen under the bearers in the photo is permeable weed
membrane. Its purpose is to stop weeds and grass growing up through the
expansion gap between the deck boards (which looks very ugly). Plastic sheet
won't do here, I think -- water will run down between the expansion gaps, and
pool on top of the pastic sheet, where it will serve as a reservoir for
evaporation onto the underside of the deck boards.
Cost of this part of project:
Pressure treated timber £200
Exterior WBP plywood £100
Wood preserver £70
Deck boards £40
Weed barrier £20
Screws, bolts, other fixings £20
Vapour membrane £10
----
£460
Skeleton and outer shell
|
|
Outer shell of the rear wall, and completed deck
|
The outer shell, including the roof, is built from 6mm exterior-grade plywood.
The wall plywood is screwed to a skeleton of 2''x2'' and 2''x4'' timbers.
I used 2''x2'' for the vertical studs, because these will be loaded only
in compression. The horizontal members are bigger, because these have to
support the weight of the roof at least partially in tension. Of course,
I could have used smaller horizontals if I had lined up the roof joists
with the studs, but I wanted the freedom to use different spacings of studs
and roof joists. In any event, this wouldn't have worked over the
door frame, etc.
Because I have no access to the rear of the building (it comes
right up to the boundary of my property) at least some of the shell and
skeleton had
to be preassembled and raised into place. I've gone to
town with the wood preserver on the rear shell plywood,
because it is inaccessible for repairs when assembled.
For some reason, it turned out that the plywood floor panels as delivered were
about about 30mm longer than the nominal 2440mm. For reasons which made sense
at the time, but now escape me (maybe I was just feeling lazy), I decided not
to cut them down to size, but install them as they were. The effect was to
leave a small area of flat timber floor extending 20mm or so all the way around
the bottom of the skeleton (the timbers for which were delivered the proper
length). Deciding not to cut the floor panels to size turned out to be a
mistake, for reasons explained below.
|
|
Start of roof construction. The large sheets of loosely-fastened plywood are to keep bits of the skeleton in place during construction
|
Choosing the roof structure was a bit tricky. Construction and maintenance
would undoubtedly
be easier if the roof were constructed to the same standard as the floor,
so two people could safely walk about on it. But since the floor cost
about three hundred quid in timber alone, I didn't really want to go to
those lengths. What's more, the floor bearers and plywood weigh
about a half a ton, and I didn't want to add that much weight to the
structure.
So rather than aim for a roof that two large people could walk about on,
I aimed for a joist size and spacing that would support one small adult
(e.g., my wife) working on crawling boards to spread the load. This gave
me a joist size of 3''x2'' at 16'' spacing (the span is about 8'6''). The
roof joists are fastened with metal hangers for speed.
The joists overhang at the front by about 18''; this is to keep rain
off the front of the building, so the doors and windows can be open even
when it is raining. This overhang does, as mentioned before, shift
the centre of mass of the roof towards the front of the building, which
has the effect of adding even more to the unequal front-to-back
weight distribution.
|
|
Most of the outer shell in place
|
Felting the roof was carried out largely by my wife, because I designed the roof
with her weight, not mine, in mind (she weighs about half as much as me).
We used the same kind of heavy polyester mineralized felt that is normally used
for flat roofs, but only the capsheet, not the underlay. It was a horrible
job installing the felt, in temperatures of about zero Celsius.
Although the bitumen adhesive we used was supposed to be cold-setting,
a certain amount of blowtorch work was still required in these low
temperatures. Oh, well; with luck the roof will be good for ten years, so
we won't have to worry about it for a while.
With hindsight, the 2'' roof overhang at the sides and back is clearly
not enough. It would be difficult to make it much larger, because
the boundaries of the property are in the way. Rain tends to run down the
side of the building, and pool on the remnant of the base that extends
beyond the skeleton of the building. As mentioned before, allowing the base to
remain slightly larger than the building was a mistake.
-- it should have
been slightly smaller if anything, so that rain could run down
the sides and straight onto the ground. OK, so a thick bead of frame
sealant stops water seeping under the floor timbers, but it's still
a bad arrangement.
Given that the side walls of the building are not clad with anything,
and there are joins in the plywood panels that comprise the walls,
if rain lashed against the walls it probably could have the potential
to seep through the small gaps between the plywood and the studs. Of course
I've closed up these gaps with sealant, but the main protection against
this problem is that fact that the boundary fence of the property
comes almost up to the walls of the building, so rain doesn't really blow
against the walls that much.
|
|
Windows in place and doorway framed
|
I obtained two double-glazed, timber framed, opening windows of about the right
size from eBay. They cost £25 each, and £25 for delivery, which
seems like a bargain. Of course, I was in a position to build the structure
around the windows, so I didn't have to look for a specific
size. I've given them each three coats of light brown wood preserver, because
I like the way they look unpainted. Because they're right under the roof
overhang, they shouldn't get too much rainwater on them. I left them
standing in place for a couple of weeks before fitting them, so with luck
they will have equalized moisture content with the environment -- if this
isn't done, there's a risk that they will swell as they absorb moisture,
which could distort them, particularly if the rest of the structure
resists movement.
|
|
Outer shell finished, hanging the doors
|
Unfortunately, I couldn't get any suitable doors from eBay, so I had to buy
some locally. These are, strictly speaking, internal doors, so I've
had to get busy with the wood preserver again. Like the windows, I've left
them standing in place for a few weeks to equalize. However, since they're
designed for internal use, most likely they're made of kiln-dried wood,
which will take much longer to reach equilibrium. I'm gusessing that
they will continue to swell for some time, so I haven't made them too close
a fit in the frame.
Like the windows, I've sized the skeleton of the building to fit the doors,
since
I'm not too fussed about the door aperture size, and planing doors to
size is a horrible job. Because the doors are only single-glazed,
as they stand they will be a major source of heat loss from the building.
I expect that it will be necessary eventually to `double glaze' the doors using
polycarbonate panels.
Fitting the double doors was a horrible job and, with hindsight, one that I
know I didn't do at all well. What I should have done was to prop the
two doors up on the doorstep with temporary supports, then built the frame
around them. Instead, I did my best to make the door frame perfectly rectangular
using a ruler and a spirit level, not appreciating that the floor
wasn't perfectly level. If the verticals of the door frame are even
a few millimetres out of square, the two doors won't meet properly
in the middle, which looks ghastly. It isn't even sufficient for the
verticals to be perfectly parallel, if they aren't square to the doorstep.
Moreover, the verticals need to be not only parallel, but in the same plane
(in the front-to-back direction).
Imperfections in the door frame can be ironed out, up to a point,
by placement of the hinges. However, more than a few millimetres can't
easily be corrected this way.
Cost of this part of project:
Pressure treated timber £160
Exterior WBP plywood £160
Roof felt £100
Doors £90
Windows £75
Bitumen adhesive £40
Wood preserver £35
Screws, bolts, other fixings £10
Joist hangers £10
----
£680
Finishing the exterior
With the shell finished, more or less, it was time to think about
finishing the exterior. Happily, three sides of the building are more-or-less
invisible, so I only had to finish the front.
|
|
Fitting the cladding and fascia boards
|
The fascia boards at the front of the roof hide the edge of the roofing
felt, and prevent it curling up in hot weather and looking ugly. They're
actually decking boards, so they match the decking at the front of
the building.
Since I was aiming for a slightly rustic look, I decided to clad the
exterior with halved logs. I'm happy enough with the result, but it
really was the most tiresome, brain-numbing job. Saw, drill, screw, saw,
drill, screw, saw, drill, screw... and, of course, nearly every
log has to be measured up and sawn to length.
The door needs a lock, of course. I've put the keyhole as high as
reasonably practicable to make it harder for the kiddies to get in.
I've also fitted hooks to the doors so they can be fastened to
the walls to keep them open in summer.
|
|
Is fitting cladding the most boring job in the world? Note that I've put the keyhole sufficiently high up that the children can't open the doors. Hehehe...
|
Cost of this part of project:
Halved logs x 30 £75
Fascia boards £10
Door bolts £10
Night latch £10
Screws and fixings £10
----
£115
The total cost of the exterior of the building, not including
the foundations but including skids, comes to £1255, which
compares reasonably well with a prefabricated shed/summerhouse of
similar construction. Of course, that's costing my labour
(probably about 200 hours) at zero £/hr. As I said, I wouldn't
expect to make a cost saving building a shed -- it's in building a garden
office that there are savings to be made.
Cladding the interior
|
|
Fitting the softwood cladding to the studs
|
For all-year-round occupancy, it is necessary to insulate the building for heat
retention and windproofing. Although a relatively thin layer of aluminium
bubble foil has good thermal resistance, it doesn't do much for windproofing
unless it is fitted very carefully. It doesn't look all that nice either, if
it's on display. For aesthetic purposes, it is better to line the inside of the
building with something, and form a cavity that can be stuffed with some form
of insulation.
I decided to form the inner wall of the cavity with 6mm softwood cladding strings because it is
relatively easy to work, and relatively light in weight. It also
looks nice, and
when it gets grubby it can be sandpapered, or washed down if varnished.
Foil-backed plasterboard
would have been a reasonable alternative -- cheaper and quick to fit.
The problem with plasterboard is that it is relatively heavy: nearly ten times as
heavy as 6mm softwood cladding for the same area. Moreover, it's not as flexible as wood, which would necessitate more or heavier studs to support it on.
The cladding I used is the cheapest tongue-and-grove variety. It is possible
to glue the strips together to form boards, and then fasten the boards
top and bottom like large panels. However, my experience is that cladding
fastened this way tends to sag out from the walls when it gets hot,
so I fastened each strip individually using the `hidden nail' technique.
This involves nailing small pins obliquely through the tongue or
edge of the grove, so it penetrates the stud at an angle. It is a
horrible, time-consuming job, and I have to wonder whether it's worth the
effort.
As well as cladding the wall, I formed the ceiling with cladding as well, and
insulated the ceiling in the same way as the walls. I did consider using 8x4
sheets of 6mm plywood for the ceiling, and I got as far as fastening a
couple of sheets up to see what they looked like. For reasons I don't quite
understand, they made the ceiling look much lower -- an effect that wasn't
apparent with cladding. Of course, plywood would have been enormously
cheaper, and taken a fraction of the time to fit. One concession I did
make was to forego the hidden-nail fastening, and just nail through
the wood. Hidden-nailing is horrible enough in ideal conditions, but on
a ceiling it's just too ghastly for words.
|
|
Stuffing the cavity with insulation
|
For the insulation I used a mixture of loft insulation and polystyrene, for
no better reason than some of each left over from a previous project.
My feeling is that, however big the cavity is, it's best to stuff it full,
to reduce the airflow path -- thermal insulation is all very well, but
a direct draught to the outside will bypass it completely.
It goes without saying that loft insulation needs to be handled with
care, especially since it will probably need to be separated into
thinner layers than it is supplied as. It's even more unpleasant to
stuff into a cavity than it is to roll out in a loft, since much more
manipulation is required.
When all the cladding was in place, I gave it two coats of thick
polyeurethane varnish. Because it was February -- cold and damp --
the varnish took nearly a week to dry to the touch, and a further
week for the fumes to stop being noxious. You can get low-odour
varnish but, in my experience, it isn't as tough as the stinky
varieties.
With hindsight, cladding and insulating the interior of the building
was by far the dullest, most time-consuming part of the job. I used
about 200 cladding strips, nearly all of which had to be measured
and cut, and then (except on the ceiling) hidden-nailed in up to seven
places to the wall studs. Each nail had to be driven meticulously at the
proper angle, using a punch to avoid damaging the cladding groves
with the hammer. It is possible to obtain concealed mounting brackets
for cladding, and they aren't all that expensive, but i didn't find
they made the job all that much easier.
Cost of this part of project:
Softwood cladding £205
Insulation £30 (estimated)
Rough sawn timber studs £20
Pins and screws £10
----
£265
Wiring
I should point out that by the time you read this, wiring a garden building
will almost certainly be covered by the egregious Part P of
the Building Regulations, so you'll need to involve the Building
Control department of the Local Authority. Of course, you're
going to want to ensure that your wiring conforms to BS7671
anyway, because it's a matter of life and death. Usual disclaimers
apply -- don't attempt this unless you're absolutely certain you
know what you're doing. If you don't know what the terms
`disconnection time' and `voltage drop' mean, hire someone who does.
The use to which a garden building
will be put will have a major influence on how you derive a supply for it.
If you'll be powering office-type
computer equipment, and maybe
a small fan heater, you won't need to do anything too arduous.
If you're installing a sauna or a pool heater more work will
be required. In general, if you can manage with a maximum current
of 13 amps, you can take a fused spur off one of the socket outlet
rings in the house. If you need more than 13 amps, you could
perhaps take an unfused spur, and put additional MCB protection
in the spur, but this is frowned on. The problem with this
approach is that increases the risk that the power ring will
be subject to continual overcurrents that are below the
level that will trip a breaker. MCBs aren't designed to
perform well in this kind of situation. For high-current applications,
most likely the safest
thing to do is to run a cable back to a spare way in
the main distribution unit for the premises.
You'll need to provide for earth
leakage protection as well, probably using an RCD, if the power outlets
in the garden building are likely to be used to supply portable appliances
outdoors. Most modern
properties have this protection already, at the main distribution unit.
If the cable run from the house to the garden building is long
(more than 40m or so), you might find that the limiting factor
on the current available to the building is the voltage drop at the
load end of the supply
cable, not the way it is supplied from the house. BS7671 dictates that
the voltage drop should be less than 4% at maxiumum load current.
For example, steel-wired armoured cable (SWA) with 2.5
mm2 conductors can nominally support 27A in an ideal
installation setup. But with a 40m run, the maximum current needs
be less than 12A to meet voltage drop requirements. Of course you
could install heavier SWA cable, but it's extremely expensive,
horrible to install, and not well matched in size to the connection
boxes you're likely to be want to use it with.
|
|
Distribution board and outlet. Note the fat SWA cable coming up through the floor
|
Consequently, as the cable run from distribution unit to the building
is nearly 40m in
total for my garden office, I decided to take my supply as a fused 13A spur.
In fact, the
maximum current that would meet voltage drop requirements with
this length of run is 12A -- less than 13A -- so to be on
the safe side I've installed in the building a small distribution unit with a 10A
MCB for the power sockets and a 2A MCB for the lights. Since the
total current available to the outlets is only 10A, there's no
point in wiring them in a ring -- a simple daisy-chain arrangement
is fine. The shed is connected to the downstairs power ring
through a fused spur unit and about 30m of SWA cable. This cable is
expensive, but it's a lot less hassle to install that digging a two-foot
trench 30m long, which is the alternative.
The fact that I'm stuck with a relatively low maximum current
makes a lot of design decisions easier. For example, I would be reluctant
to run cable in the cavity between the inner and outer shells of the
building if it had to carry a higher current. Because the cavity is
completely stuffed with insulation, in places the cable is likely to
be completely surrounded by it. In such circumstances BS7671 dictates
that the cable be derated to 50% of its nominal current-carrying
capacity. Because my maximum current is so far below the nominal capacity
of the cable, this isn't a problem for my installation. If it were,
I would have to consider installing cable in conduit on the surface
of the inner walls -- something that I really wouldn't want to do.
One thing that certainly merits careful consideration when wiring
an outbuilding is waterproofing. BS7671 doesn't have a huge amount
to say on this subject, beyond a general requirement to pay attention
to the environmental conditions, which are likely to be
rather hostile. However, so far as I can see, nothing in BS7671
requires that an outbuilding be wired as if the wiring were
outdoors. That is, I can't see any particular requirement
to use waterproof socket outlets, waterproof cable entries, armoured
cabling, etc.,
provided that all reasonable precautions have been taken to prevent
the entry of water into the electrical fittings.
As it happens, I have used splashproof (IP44) lightswitches
and socket outlets, not because I'm worried about rainwater
getting into the fittings, but because I'm aware that they are
likely to be handled by people with damp hands. I've run
ordinary flat twin-and-earth cable in the wall cavity;
with tight-fitting gromets and a big blob of sealant at
the cable entry point, I believe that water entry is effectively
prevented in all normal situations. I wouldn't risk flush-mounting
socket outlets, because I don't think an ordinary flush-mount patress
box can be made sufficiently waterproof. I should also point out that
I've built my office for year-round occupancy, and I'd know very
quickly if water was entering the part of the structure where the
wiring is.
If you run the wiring in the wall cavity,
it may not be a terrifically good idea to put junction
boxes in the cavity (even if waterproofing is not an issue, which
it most likely will be) because they will be difficult to get to
for maintenance.
Please note that if you've got extraneous conductive parts -- copper
pipes, structural metalwork, etc. -- you'll need to pay some atttention
to maintaining equipotentiality between the metalwork and the supply earth.
This subject is beyond the scope of this article.
SWA cable £50 (50m drum, not all used)
Flourescent luminaire x 2 £40
IP44 Socket outlets x 4 £40
Small consumer unit £15
MCBs £15
Flat twin-and-earth cable £10
IP44 lightswitch £10
Cable fixings £10
13A Spur unit £5
----
£195
Flooring
There really isn't an inexpensive way to make a nice-looking floor,
which is washable and hard-wearing enough for a garden building.
I considered and rejected carpet (washable types way too expensive), vinyl
tiles (inexpensive ones too ugly), coconut matting (difficult to
clean), painted plywood (difficult to clean and ugly), and a
number of whackier alternatives that I'm too embarassed to mention.
In the end I decided to use laminated wooden flooring, the cheapest
I could get. This turned out to be Wickes' own brand 7mm beech-effect
glueless flooring, at a whisker under £7 per square meter.
The only problem
with this stuff is that I don't expect it to be particularly hard-wearing --
the wood layer is tissue-paper thin, on top of a substrate of MDF.
But it seems to be doing OK so far.
Because the plywood main subfloor was not particularly even, I decided to
use underlay under the main floor, and also a vapour membrane to prevent any
moisture that manages to ooze under the walls from dampening the main
floor. The edge of the vapour membrane is hidden under the skirting
board, so the membrane forms a `tub' shape, with the laminated
flooring and underlay in the bottom of the tub.
Laminated flooring cannot be laid right to the edge of the floor, because
allowance has to be made for expansion. You need wedges (or
offcuts of wood, which are cheaper) to stand the floor panels off a uniform
distance from the walls. Glueless flooring needs to be hammered together,
which is difficult at the very edges where there isn't clearance to
swing a hammer. There's a specific tool for dealing with this situation
-- a bit of bent steel which hooks over the edge of the flooring and
provides a sticky-up plate to hit with the hammer. This tool usually comes
in an installation kit that includes the wedges and a hammering block.
You'll need the hammering block, because if you hit the edge of cheap
laminated flooring with the hammer itself, it will collapse.
Cost of this part of project:
Laminated flooring £75
Underlay £20
Installation kit £10
Vapour membrane £10
----
£115
Finishing and trim
|
|
Windows are finished with architrave; skirting, bevel, and quadrant also visible
|
By this stage I had the exterior finished and the interior clad, wired,
and floored. The whole thing was starting to look like a garden office, rather
than a shed. However, there were lots of unsightly gaps between the edges
of the cladding boards, and some of the structural timbers were still
visible. What's more, there was an ugly expansion gap around the floor,
and the vapour membrane was still visible. Both problems were fixed by
installing skirting board. In fact, to save the expense of full-size
skirting board, I used architrave. I could have used quadrant, I guess --
the gap was only about 8 mm. But quadrant isn't hugely cheaper, and
architrave looks like real skirting, albeit smaller. I closed the
corner gaps with quadrant beading, and concealed the joins between lengths
of cladding using bevel beading. Both types are surprisingly expensive. The
visble structural timbers I boxed in and clad with planed stripwood.
I could, I suppose, have used the leftover walling cladding, but it would all
have had to be cut to width, which would be a drag with the kind
of tools I have (I refuse to use a bench-mounted circular saw, because
I need all my fingers to type with).
Cost of this part of project:
Skirting board £40
Quadrant £52
Bevel £12
Stripwood £50
----
£154
The total cost of the interior, including wall and ceiling cladding,
insulation, flooring, electrical fittings, lights, and trim was
£729.
What I learned
OK, so the following may be obvious to some people, but even after ten years of
hardcore DIY activity, including rennovating a delapidated
house, they weren't obvious
to me.
Wood preserver is designed to soak right into wood. Obviously --
it wouldn't work otherwise. What this means is that if you spill coloured
wood preserver on wood where it isn't supposed to go, you'll
never get the colour out, even with sandpaper. I used green
wood preserver for the parts of the building that wouldn't be clad, and which
might just about be visible. But it's very obtrusive if you spill it
on wood that is intended to remain wood-coloured.
Bitumen dissolves in wood preserver solvent. In places where
bitumenised roofing
felt came into contact with timber that has recently been treated with
wood preserver, the wood preserver dissolved the bitumen and soaked
it up into the wood, leaving an ugly black stain that (naturally)
permeates the entire thickness of the timber. Make sure wood that has
been treated is fully dry before putting it near roofing materials.
You can't screw two pieces of wood together so tightly
that water cannot penetrate. Because wood, particularly unplaned
wood, is never perfectly flat, the join between timbers will always
be big enough to let some water through if it gets wet. There needs
to be sufficient access to get some frame sealant around the join
after construction.
Water needs to drain away. If you allow water to pool
anywhere, it will most likely find a way into the building.
It's impossible to make a double door frame sufficiently
accurate by measurement alone. The frame needs to be built
around the doors. Building a double door frame is completely
different from building a single door frame -- with a single
door the battening around the door will hide quite substantial
inaccuracies in the frame or the hanging. With a double door,
it's all on display.
Mixing concrete in large quantities is hard work. If you have to make a
foundation, even in the form of concrete islands like mine,
it's got to be worth hiring a mechanical mixer. I couldn't,
because it wouldn't fit through the house.
It takes much longer than you might think. It isn't the substantial
construction that takes all the time, its (a) the preparation, and
(b) the fiddly little jobs that you forget to account for. Preparation
includes things like treating timber with wood preserver (it's slow,
tiresome, and it takes ages to dry), digging and levelling ground,
and carrying and stacking huge piles of timber. Fiddly little jobs include
fitting door furniture, trimming and finishing roof felt, picking all the
stupid barcode labels off everything (I just hate that), sealing
up gaps between timbers. If you can't work on the job continuously,
you also have to account for the time consumed by getting ready to
work and tidying up, and the constant sweeping, cleaning, and tidying
that are required to prevent the site turning into an eyesore.
The quicker the roof goes up, the sooner you can use the building for storage. If you're working intermittently, it's much easier to be able to leave
tools and materials in the partially-constructed building, than to get them
out and put them away every time you want to work. But most tools and
materials aren't particularly waterproof, so you'll need to have at least
one wall and part of the roof up before you can leave stuff in the
work area even overnight. If you have even a few roof joists up, you
can make a temporary roof of (say) corrugated plastic sheeting, even if
you're going to use something different for final roof. The time you'll
save being able to leave tools and materials in the building more than
compensates for the time taken to put up the temporary roof. But don't be
tempted to put up part of a plywood roof -- ideally it all needs to go
up together and be felted the same day. Roofing felt doesn't stick
very well to wet plywood.
October is a bad time to start work on a major outdoor DIY project.
It gets dark early, it rains all the time, and it's cold. And if the work takes
longer than you anticipated, you'll be trying to felt the roof in January
when it's freezing.
Some jobs are too big to do single-handed. However enthusiastic,
strong, or competent you are, some jobs need more than one person -- it's
as simple as that. Screwing the fascia boards to the overhanging roof
rafters was a case in point. With one person to hold one end up,
and another to hold the other end and screw the boards to the
rafters, it would have been a five minute job. Single-handed it was
a two-hour job. Longer, if you count the time I spent dressing he
wound after I screwed my finger to one of the rafters.
Yale-type night latches are designed for inward-opening doors.
Shed doors typically open outwards, because there's more room
outside than inside. If you fit an inward-opening latch on
an outward opening door, you'll find that you need a key to
close the door, but can open it without one. This, of
course, is not ideal. I was able to dismantle the latch I bought
and reverse the locking tang, but it still means that when
you close the door, the locking tang strikes the wood, not the
metal of the lock body. I'm not sure if anybody makes a
night latch specifically for outward-opening doors.
Conclusion -- how much did I save?
With hindsight, building the Kabin turned out to be a more
ambitious undertaking than I really anticipated. What's more,
I deliberately did the work during the winter, because I had
other plans for the summer, and the days are rather short in
December. I started work at the start of October 2004, and finally
declared the job done at the end of March 2005. That's seven
months from start to finish. My recollection is that I worked
for three full weekends (7am Saturday to 6pm Sunday), a couple
of Saturdays, and between 30 minutes and an hour about
three evenings a week, with a break of three weeks over
Christmas). This adds up to about 300 hours of work. A team of three people, working a typical 8-hour day,
would have racked up these hours in about three weeks. Working
alone, evenings and occasional weekends, comes to seven
months.
In fairness,
when working out the cost saving over buying a commercial garden office,
I really ought to account for the time I spent driving back and forth
to timber yards and builders' merchants,
searching the Web for suppliers, and poring over my computer doing
the design and calculations. But I can't -- I really can't remember
how long all this took. On the other hand, if I had opted for a commercial
offering, I guess I'd have spent a
considerable amount of time investigating manufacturers and comparing
prices, so maybe it balances out.
As for the financial cost:
Foundation £200
Frame and exterior £1255
Interior, electrics, and trim £729
----
Grand total £2184
(This price does not include the desk and shelving that can be
seen in some of the photos -- these were leftovers from another
job).
For the purposes of comparison with a commercial garden office,
the cost of the foundations should be excluded, since foundations
would be required in both cases. So the relevant figure is
1984.
It's hard to be sure how this bottom-line figure compares with
a commercial garden office. Henly's MidiSpace has a similar
internal size, and similar features, and costs about
£12000 (assembled). But it doesn't have
solid wood walls, floors, and ceiling (they're glassfibre and
MDF). This price is, if anything, towards the cheaper end
of the market.
So, by building my own, I estimate that I have saved at least
£10000 -- a figure not to be sniffed at. Allowing 300 hours
labour, this makes my `earnings' £33 per hour. All in all,
not a bad deal.
©1994-2006 Kevin Boone, all rights reserved
