Garden studio

Here’s a detailed description of the building of a hempcrete studio in my garden, which I started last summer. I have completed the base, and the hempcrete walls and roof and will soon be posting about how I did the floor.

The base is an unusual foundation, but it will keep the hempcrete building dry, solid and safe. I began by digging a trench around 3 inches (75mm) wide and 7 inches (175mm) deep in the shape of the building footprint. I then poured concrete into the trench and 2 inches (50mm) above it with some formwork. There is one thin rebar in the concrete to aid tensile strength. When the ring beam was set I dug out inside to a depth of around 6 inches (150mm).

A geotextile membrane (Terram) was then laid over the ground and the base was filled with Glapor foamed glass insulation. Glapor is made from recycled glass in Germany. The foamed glass was then consolidated with a vibrating plate and another layer of geotextile was added.

The rationale behind this base is that it enables a garden building to sit very near to the finished ground level, whilst remaining dry and insulated. The Glapor is an amazingly good product as it will not wick water, meaning everything above it will stay dry without the use of a plastic membrane. If in the future the building were to be dismantled the Glapor could be reclaimed, and the concrete ring beam is a minimal amount of concrete. The Terram is necessary as it acts as a filter to stop the ingress of dirt that could be washed in over time or through flooding. If enough dirt did wash in, it could allow water to wick and would reduce the insulation efficiency. The combination of foamed glass and ring beam will give plenty of load-bearing in this local soil with what is effectively a solid, but relatively lightweight, building. Pebbles could be used in the same way but would lack the insulation quality of the Glapor.

A damp-proof membrane isolates the timber frame from the ring beam and the framing was built to follow the ring beam. The frame was built up from the base in 3 x 2 (75 x 50) timber and the walls will be around 8 inches (200 mm) thick. This is a very appropriate size as it protects the timber and allows the hempcrete a reasonable covering over the timbers on both sides. It will also provide excellent comfort levels for the UK climate. The floor joists will be infilled with 3 inches (75mm) of hempcrete and a chestnut timber floor will be laid on top. This will create a very well insulated dry floor.

The Geoplast shuttering is really quick and efficient to use and I would advise any serious hempcrete contractor to get some, although it is very expensive! I used 4mm ply for the curvy walls as I had some left from another job.

I made the timber frame of the base curved largely in the shape that I wanted it and then just banged in some battens into the ground. Then I bent some 4mm ply to the form I wanted and screwed it to the frame and the battens to hold its shape.

This then gave me the form for the outside wall shape. Then I took another sheet of 4mm ply and cut it in half lengthways to make two lengths 2ft high (600mm). I joined them together and fixed them to the floor base 8 inches (200mm) from the outside shutter. As they are curved they largely support themselves, but I added enough support to hold the two shutters in place. For this kind of building I am not trying to get perfect smooth lines so the fact that the shuttering is not perfect just adds to the interest of the walls.

We then started hempcreting and placed the first layer all the way around the building to form the base of the wall.

It took three of us about two and a half hours to hand-cast around the building for one 2 ft (600mm) lift with one person mixing and two hand-placing. The shuttering was removed the next day and the wall scratched back with a nail float where necessary to form the shape of the wall I wanted. When I’m using curved plywood shuttering once I am happy with the shape I give the wall a couple of days to get hard and then I simply screw the next layer of shuttering ply into the hardened hempcrete with 5 inch (125mm) screws. This holds the ply solid enough to hand place the next layer.

This photo shows the first two lifts in place and the framing of the shuttering for the windows goes in at this point. The timbers fixed to the wall plates at roof height cross the building in various directions and so brace the timber frame to stop any movement whilst the hempcrete sets.

The two mixers you see in this photo are the ones I use for small garden jobs. The binder and water are pre-mixed in the small drum mixer. The hemp is placed in the pan mixer on the left and switched on. We take off about a bucket and half of slurry from the drum mixer and pour it into the pan mixer as it turns. This has two enormous advantages. The first is that it makes the whole process much less dusty as it only dusts for a few seconds in the drum before a spray from the hose calms the dust. The pan mixer creates hardly any dust with this method and makes the mixing much more pleasant. The second big advantage is the speed of mixing. A batch of hempcrete will mix in about 90 seconds this way as opposed to 4 minutes when hemp, binder and water are mixed separately. This means that even though we are mixing relatively small batches at a time, we can mix and place up to about 5 cubic metres of hempcrete a day with three of us.

We carried on raising the straight walls using the Geoplast shuttering. Straight walls like this are much easier to shutter, but I really enjoy creating the curves.

Here’s a very simple way to create curved corners. I staple render mesh to two wooden battens and fix the battens to the frame each side of the corner.

As we put the hempcrete in we push a little harder to stretch the render mesh into a curve.

I then use a nail float to scratch the finished corner to the shape I want. You can make a nail float quite easily with a piece of flat wood with a handle and screws sticking out of the face.

Here you can clearly see the three lifts and a rough edge. This is because I would normally have built all the framing for the door and windows and know what sizes I wanted to do. The truth is I haven’t decided what size of windows or door I want to make or salvage so I just left a larger opening. When I decide on the windows and doors I will frame it up and spray them in to make good.

About half-way there! Notice the far end of the wall doesn’t have any timber framing in it. Hempcrete is a remarkably robust material and in these small garden buildings missing a bit of framing won’t make any difference to anything.

If you notice the small level on the top of the shutters it’s because I level the first shutter horizontally so that when I attach more shutters along the wall the line stays parallel to the top of the hempcrete lift. If you put it on at an angle the shutters start to run out of line with the top of the previous lift.

It is at this point that having trestles or a small scaffold tower are really useful.

At the top of the curvy wall now. I have shuttered up to just above the wall plate so I can scratch back to the wall plate to form a straight edge. This will make positioning the roof timbers easier.

The rose bush will eventually grow over the roof, but for now it has been pulled back away from the roof to make the hempcreting easier.

This end wall has a small window in it so I made a simple formwork to give the hempcrete roughly the right shape. I usually make the form a little smaller than the window and scratch back the hempcrete to create a tight fit. The diagonal brace will keep the opening plumb and square. There is no framing or lintel in the hempcrete here as the material is plenty strong enough to cope with this opening.

Using a full sheet of ply like this and simply screwing it to the hardened hempcrete makes the shuttering follow the line of the existing wall exactly. The curve adds solidity to the shuttering meaning no extra support is required

And finally the top lift! We usually do it so the person mixing the hempcrete feeds the person on the scaffold so there is less getting up and down for the one who is hand-placing.

The cross timbers at roof height were removed as the frame is firmly set in the hempcrete now. There is no diagonal bracing on this building as its shape and small scale make this unnecessary.

The curve emerges from the shutters! You can see at the top of the wall two sections of wall plate showing in the hempcrete. This is where the framing is in two short straight sections of studwork. These two sections of studwork are joined at the base and the wall plate, but no attempt to build a curved frame has been made. Two short straight sections of studwork across the two curves gives all the load-bearing strength the roof will ever need and this avoids having to form complicated curving frames.

The rounded corner is revealed and behind that, the wall bulges out around the window. This started out as just some overenthusiastic hand-placing pushing the shutter out, but I liked the shape it left so I carried it on up the wall. When the rosebush grows up over the building the window shape will be more visible.

All the outside shuttering is removed and the walls are largely finished.

The roof structure here is a simple flat roof made from 150mm by 50mm (6” x 2”) timbers spanning between the wall plates. There is a small fall across the roof to help the water run-off. I would have liked to have built a pitched roof but UK planning laws prevent a garden studio from being any higher than 2.5m because it is against the boundary of my garden.

The rafters are noggined in two places to keep them level, straight and equidistant. Half thickness noggins are used to allow continuity in the hempcrete at ceiling level.

This is a view from underneath after I had fitted a temporary roof for the winter. Note the wood wool board fixed with washers and stainless steel screws to the underside of the timbers. Stainless steel screws are required here because the screw heads will be directly below the lime plaster. Ordinary metal coated screws can deteriorate over time and rust, creating red spots on the surface of the plaster.

The hempcrete will be placed in between the rafters and come down to the face of the wood wool board. I could have stopped the hempcrete at the bottom of the rafters and not used the wood wool board, but this would have two major drawbacks. It makes the plastering more difficult as the plaster draws differently on the timber and the hempcrete. This means the plaster would be ready to trowel up on the hempcrete before the plaster over the timber is ready. This can lead to dragging in the plaster and makes finishing the plaster more difficult. Using the wood wool board makes this easier. The second and perhaps more important reason to use the wood wool board is because the adhesion of the plaster to the timber over time is not that good, as timber will move differently to the plaster in relation to moisture. This can lead to the plaster coming away and cracking in lines at the timber. Having timber in the surface of hempcrete generally means a mesh is required in the plaster to reduce the risk of cracking. Mesh introduces another non-natural material and is not necessary with good design. Mesh will largely prevent cracking, but it will not stop the plaster and the timber separating over time and leaving regular weak points at every timber. You can see in the photo above that the first few bits of shuttering are in place.

These are a couple of views from above showing the rafters with the shuttering now in place underneath. Note the timber battens (38 x 25mm 1½” x 1”) on the sides of the rafters. These battens act as hooks to prevent the hempcrete from falling or slumping down over time. They will also reduce air transfer if there is any shrinkage in the hempcrete.

We sometimes use this technique on the side of timber studs in walls too, where the frame is on one side of the hempcrete. This is to lock the hempcrete into the frame, preventing movement and again reducing air transfer.

When hempcrete is hand-cast into individual panels between timbers with no overlap the risk is always that it shrinks away from the timbers as both timber and hempcrete dry. Spray-applied hempcrete does not shrink back, as the adhesion to the timber is greater than its potential to shrink through drying. Shrinkage can be a problem as it is obvious that there is no point having a great insulative material if you are just going to let air flow around it, carrying heat away as it goes. The best ways to avoid this issue when hand casting is to ensure the timber is dry and use a slightly drier hempcrete mix than usual. The drier mix reduces shrinkage.

In this photo you can see a drum mixer for premixing the binder and water, and a pan mixer for mixing this slurry with the hemp. This technique is ideal for making a dry hempcrete mix as you can control the binder and water content easily for each mix.

Small gaps in the shuttering (below 15mm or ½”) are not important as the hempcrete will block the gap. When the shuttering is removed the slight line of excess hempcrete can just be scraped off.

Scaffold boards are used to make walking on the roof easy and a piece of timber is used to rule off the hempcrete. Small pieces of scrap plywood were used to close off the ends between the rafters. These were fixed onto a batten running across the rafter tops to hold them in place.

When placing the hempcrete make sure that the hempcrete is pushed well under the battens on the sides of the rafters. Keep the hempcrete layers thick as you do not want it to delaminate from underneath once dry. Placing a thin layer of hempcrete onto the shuttering and putting a second layer on top makes delaminating more likely.

I was concerned on the day we hempcreted that it might rain so I moved the plastic around on the roof to gain access for the hempcrete, uncovering half the roof at a time. This meant the plastic could easily be pulled over quickly if it had begun to rain. I still enjoy the look of a well-filled hempcrete roof!

I hempcreted this roof in January and the temperature on the day was adequate, but the weather was cold for the following three or four days. I used the driest mix I have ever used with the least amount of binder I thought I could get away with. I did wonder whether the roof would work or if I would have to pull the hempcrete back out again under these conditions. I left the shutters on for several days to give it the best chance and watched as the hempcrete gradually got harder. I am pleased to say that in spite of pushing the limits of the material it has set very well with no shrinkage at all.

In the bottom left of the picture you can see how the simple shuttering is bulging a little on the roof edge. This can be just scratched flush with the wall tops when the hempcrete is set. Sometimes sloppy shuttering and a scratch back is quicker than perfect, but time-consuming, shuttering.

The rear wall of the building against the fence was already plastered as I had limited access to the wall, so the edge of the roof section had to be plastered up to the top before I could put the final layers of roof on.

Once the plastering was complete on the back wall a breathable membrane was laid down to cover the hempcrete.

This layer is important as it serves two purposes. It closes the surface of the hempcrete down to avoid wind robbery. Wind robbery is when air is forced across and into the insulation layer, drawing out heat as it goes and reducing performance. The second function of the membrane is to protect the hempcrete from any condensation drips from the underside of the roof covering.

This photo shows the membrane held down with 10mm (3/8”) thick Douglas fir spacers.

These spacers hold the OSB roof board off the hempcrete and provide an air gap for air to flow across the roof. By using short lengths of timber to pick up two thirds of the OSB boards good air flow to cross-ventilate the roof can be achieved. If the spacers were long lengths, continuous down the rafters, the air could only flow in one direction.

Leaving the airspace between the hempcrete and the roof covering open to insects could cause problems later on with unwanted guests so the perimeter of the roof was closed down to stop insects, but allow air flow. Rather than buy the standard insect mesh and try to fit it I decided to make my own by rolling 5mm render mesh up into a roll and holding it with tape. Each roll was cut to length and was thick enough to fill the space to hold itself in place. The air flows easily through, but the holes are too small for insects to pass.

The OSB board was then screwed to the rafters through the spacers to create a roof deck for the final waterproof roof coating. I am growing a large rose over the roof so I used a fibreglass roof to protect the building and withstand the abrasions from rose thorns.

Here you can see the ceiling from underneath with wood wool board covering the rafters to make plastering easier and more reliable over time.

Here’s how I dealt with the boundary being so close to the new wall of the garden studio. The rear wall of the studio backs on to my neighbour’s fence and so access for the build and in the coming years will be very limited. I had several options of how to deal with this, but the one that appealed to me the most in this case was to plaster each lift as I went. This was an opportunity to test the viability of the idea, but also to look at the potential issue of the tannins leaching out through the lime plaster if it is applied straight after placing the hemp.

To hempcrete the rear wall I made up a row of shutters the full length of the building and attached ropes to overhanging timbers. This way I was able to hang the shutters at the desired height all the way up the wall. I didn’t have access to the back wall to screw the shutters to the frame so I used wires through the shutters to pull them towards the frame. I simply pulled the wires out through the hempcrete to remove the shutters.

I set the shutters up for the first lift and hempcreted them. I removed the shutters and lifted them up to the top on the ropes so they were out of the way. When they were tied off I mixed up some plaster and leaned over the wall to plaster the first lift. The following day I set the shutters for the second lift and hempcreted it. When the shutters were removed, I dropped them down to the floor and plastered the second lift. This process was repeated to the top of the wall.

Plastering upside down onto fresh hempcrete was easier than I thought it would be. I knew I would not get a cosmetically impressive finish because I am not a plasterer, but I am confident that the plaster will protect the hempcrete wall for a very long time to come. The fact that the plastering is not well done from a cosmetic point of view is unimportant as the wall cannot really be seen from anywhere.

difficult plastering

This method was very cheap and easy on materials and I rather enjoyed the challenge, so I think I would be happy to do it again another time when dealing with a boundary fence or wall. The wall will get little sunlight and almost no rain as the roof will overhang to the boundary so the wall should sit there unchanged for the foreseeable future.

how to plaster in a tight space

Interestingly you can’t plaster using a hawk when you plaster upside down as the plaster just falls off, so I just screwed some plywood to the wall plate and used a second trowel to cut the plaster off the spot board, moving the board as I went. I left the plaster short of the ends of the wall as I will wrap the end of the rest of the plastering around the back of the wall so the joint is not visible from anywhere. There has been no tannin staining on the plaster which I find interesting as some may have predicted that it would stain being plastered onto wet hempcrete.