This is a post I have been meaning to write for a while: a review of the experience of clay plastering. The first point to make is that I don’t think I arrived at the absolute definitive method, finding myself constantly experimenting with different mixes and application techniques in order to find the optimum balance of the various considerations involved: these being adhesion, cracking, gap filling and time.
It was much harder to get clay to stick to straw than the Limecote I used on the outside which could be sprayed-on and built-out to possibly even 40mm in a single application. Using clay, it is necessary to put-on an adhesion coat. This should be mostly pure clay, to make it sticky possibly with a little sand added to give it a bit more strength. This layer can’t be too thick, because a clay rich mix, while sticky, will crack a lot and also will not be strong enough to allow it to be built-out. It s also worth noting that clay is more expensive than sand, so you don’t want to use more than is necessary. The issues with this adhesion coat are how sloppy to make it, how thick to apply it and how to apply it. My solution was to make a pretty sloppy mix (consistency of double cream) and apply it with a gun to a thickness where the straw was coated and stray sticky-out bits could be encouraged to lie flat but not to the point where the coat starts to build-out.
I did find that I got some areas where, after applying subsequent build coats, the clay was not strongly bonded to the straw and therefore was inclined to crack, but this mostly occurred where there was weakness with the underlying straw surface, either because the straw was rather soft and wispy or because there was a gap which wasn’t adequately packed-out. Where these were an issue I hacked off the clay and started again, making sure the clay was very well worked into the surface.
I probably wasn’t as diligent as I should have been on gap filling. This was partly because the Limecote was so good at bridging across gaps or soft areas and also because of the time required to do a thorough job packing out gaps with a straw mix. I therefore tended to rely on my first build coat to fill-in any gaps but this could lead to cracking and weakness if I didn’t sufficiently force this coat into the gaps. For this reason it would probably be best to spend some time really packing-out any gaps with a long straw dubbing mix.
Cracking can happen for two reasons: either your mix doesn’t have enough sand and/or fibre added to it, or if the clay is insufficiently bonded to the straw underneath – because the straw surface is wispy and unstable or there is a gap beneath the clay that isn’t well packed-out.
I think I got the fibre issue cracked (as it were) by using concrete screeding fibre. This is much easier than using chopped straw because you don’t have to chop it. I did use chopped straw initially and found the best way of prepping this was to put a load into a dustbin and then plunge the strimmer into it, preferably with someone holding the dustbin lid over the top to prevent too much flying out. The only issue with fibre, especially screeding fibre, is that if you use a sponge to finish-off your top coat, this will drag bits of the fibre out, resulting in a hairy finish – but this can be dealt with by using a blow torch to burn these off.
As mentioned earlier, I got some minimal cracking of my first build coat, but this was mostly over gaps. I dealt with this by tapping the cracks with a hammer, and if the area around the gap sounded hollow, I would knock the gap out and ram clay (or sometimes dubbing straw) into it. If the gap seemed ‘natural’ and reasonably firm either side I just left it and filled it out with the second (and final) coat.
I didn’t get any cracking of my final coat – which I figured must have meant that the mix was right and the fibre was doing its job.
I did initially experiment using a fibreglass mesh which I stuck on top of my first build coat. This did stop any cracking but I was deterred from using it extensively because the supplier of it said it should be sandwiched into the final coat which effectively meant making the final coat two coats (as in two bits of bread if the mesh is the butter in the middle) and this was too time consuming. That said using it as a layer between the first and second coat didn’t seem to create any problems, such as creating a weakness or potential for ‘shearing’ between the two coats – although I guess only time will tell if this might develop as a problem.
I underestimated the length of time it took to do the clay plastering, possibly because the experience of the external rendering was easier than expected (Limecote being such a magical material to work with) or perhaps because the internal walls meant there was a considerably greater area to deal with. Using the gun didn’t seem to speed things up, other than when using it to spray the adhesion coat. When I used the gun for the first build coat, I found that I could go faster by hand, especially if I kept this layer relatively thin and when it came to the final layer pushing it on with a trowel was also faster than standing there with the gun waiting for it to build-up to a sufficiently thick layer. This was probably because I was using a basic gun and compressor. If I had more powerful kit I probably could have sprayed thick coats on quite quickly.
Because of the length of time things were taking I also tended to miss-out the gap filling stage, as mentioned above. This was probably not ideal, but I don’t think it has caused any long-term problems (i.e. clay dropping off at some point in the future), but I guess only time will tell on this one.
If time (or labour) wasn’t an issue the absolute best process would probably be the following.
1 Adhesion coat
Apply a slip coat of 3 to 1 clay to (plastering) sand with a gun, mixed to a consistency of double cream. Apply this to a thickness such that the straw is well coated, but not starting to fill-out.
2 Gap filling
Make up a batch of long straw dubbing mix. This is straw with a clay / sand slurry poured into it. You could make this slurry by taking the slip coat and adding some sand to it to firm it up a little. This mix should have the characteristics of straw bound together with clay, rather than clay with straw added to it. Use this to pack out any gaps or soft areas, ramming it in as firmly as possible. Unlike the lime (a chalk fibre lime such as Limecote), clay is not very good at bridging across gaps and soft areas and even if it appears that you have covered a soft areas well, the likelihood is that it will be weak and crack easily as it dries. You could (probably should) do this while the slip coat is still wet. This should get you to the point where the surface is relatively flat and firm, but still basically strawy.
First build coat
Make up a batch of clay that is 3:1 sand to clay with fibre added. I used concrete screening fibre (a couple of handfuls per 60kg batch – ie one 15kg bag of clay powder plus 45kg of sand). I mixed this in a regular cement mixer, tumbling the dry mix first and then adding sufficient water to take it beyond the golf ball clumps stage, past the big lumps stage to the point where it starts to slump. Something like the consistency of cream cheese. Then apply this by hand, really ‘grinding’ the mix into the straw, using the heel of your palm, having first dampened down the straw with either a mister on a hose or by using a pump-action sprayer. This layer need not be that thick, but you should get to the point where the wall no longer looks strawy, but more clay like, with odd bits of straw poking out.
Let this coat dry out so that any cracks can fully develop. Then test the cracks by tapping them with a hammer to see if they sound hollow around then. Those that are hollow, knock-out and fill with a clay mix or possibly by ramming a long straw mix into the gap. Alternatively you could wait until you start to see cracks develop as the coat dries and then push the still damp clay into the straw along the crack line with your finger tips.
Second build / final coat
Get the wall good and wet with a mister. It is amazing how much water the dry clay wall will suck-up. I tended to give it a good soak, to the point where water is just about to run-off the surface, and then wait five minutes or so and give it another go. Apply a 3:1 sand to clay mix with fibre using a trowel to a sufficient thickness to give a decent covering to any straw poking out. I find descriptions of coats by thickness rather unhelpful with straw, since its irregularity will mean that a coat which in one place may just cover the straw, will in others be 30mm thick. I used the trowel to get the surface reasonably flat and smooth, but tried not to work it too much, and then used a wet sponge trowel to further smooth over the surface. This produced what I guess you would call a rustic finish, which I was happy with. If you wanted a better finish you would probably need to do a third skim coat which you could spend any length of time working-up to whatever finish you want, rubbing it with a wooden float or a stone or whatever. The great thing about clay is that it doesn’t set, it simply dries-out so you can rework or revisit the surface as much as you like.
Finally, once this coat is dry, get a blow torch and ‘wash’ the surface with flame to burn-off any bits of fibre or straw.
It has been a long time since the last post – mostly because we have been doing ‘non-straw’ jobs, such as installing bathrooms and kitchens and painting which are not especially interesting or noteworthy tasks. I have been meaning to write a review post covering my conclusions from clay plastering, but haven’t got round to it yet. However, last week I got back onto a straw specific job, namely finishing off the exterior render around the windows.
I had left the window area without a finish coat because I needed to get the windows themselves in before doing this and I wasn’t able to do that job before the winter lime rendering window closed. As luck would have it I received a surprise visit from Arthur Phillip, the lime render man, who happened to be in the area. Arthur is probably the country’s foremost expert on rendering straw bale houses and he introduced me to PVC window frameseal beads. These are basically a strip of white PVC channel with a sticky back to it that you can put around the wooden frame edge of the windows allowing you to render up to and into the channel – thus giving you a stable and watertight edge to your render. They also have on the a tear-off sticky plastic strip which you can use to attach polythene sheeting to in order to protect the windows themselves from render spalshed – important if you would be using a gun to apply the render.
I couldn’t actually source these in sufficiently small quantities, so I made my own by buying standard PVC channeling and runing a bead of silicone up the back of them to stick these to the frame while also creating a weathertight seal. I then mixed up a batch of lime and straw to create the shaping into the window frame and finished-off with a coat of lime that then ran into / up to the window edge fo the PVC channel. This seems to work fine.
I am now getting round to the fiddly detail bits. One of these is fitting notched bales between the top the main interior wall and the roof (the bales have to be notched to allow for a purlin) and also putting bales in place above the doorways. It was good to return, one last time to ‘straw work’ athough cutting the notch with a chainsaw meant becoming swamped in loose straw. It also reinforced that by far and away the best tool for cutting or shaping straw is an electric chainsaw (not a petrol one because the straw gets in the air intake and it overheats).
I have also been finishing off the clay around the windows and door frames – using a ‘cobby’ mixture of straw and clay. I want to plaster right up to the window frames, rather than put decorative wooden frames and window ledges in place. I think this sets-off the window frames to best effect, but it does mean plastering over the window sub-frames as well as sticking clay ‘upside down’ to the sub-frame above the windows. To do this I covered the wood in a flour, water and clay paste to give it maximum adhesion. This appears to have worked – although for the final coat I will also stretch hessian or fibre-glass mesh around curve, both to stop cracks developing over the join and also to hold the upside down plaster in place should it feel like falling-off. The job feels a little like being a housemartin, sticking clay and straw to the underside of eaves – and the resilience of their nests gives you some confidence that the clay, with added flour paste, will be sticky enough.
We have now had a chance to see what happens as the base coat dries-out. There are some cracks developing – but not many. I have probed any large crack to see if this indicates any fundamental weakness in the straw or adhesion beneath, and for those where there is significant softness I have pushed my fingers in more strongly and forced the clay back into the straw / gap – and then filled this out with more base coat.
I have also changed the approach to the first coat – moving away from a very liquid pure clay slip, to a mix that is 1:2 clay to plastering sand (no fibre). This gets over the probem of pebbles blocking up the gun because the sand is finer and better sieved (albeit more expensive). I get the mix to the point where it is a cross-over between liquid to solid but where if you dip a piece of straw into it you still get a good amount clinging to it. I have then sprayed this on just to the point where you stop just covering each bit of straw and start getting ‘fill’ between the bits. This we will then leave to dry-out a bit before putting the 1:3 base coat (plus fibre) on.
I have also noticed there is a marked difference between the bales whch present a cut face and than those with a folded straw face. The cut face is already pretty rigid and stable and therefore doesn’t require much to cover it, whereas the folded face is more wispy and requires more material to really get into the straw and stabilise it.
This past week or so I have been getting on with the second coat, which I also intend to be the finish coat. The mix I have been using is firmer than that which went through the gun for the first coat, but still relatively sloppy – given the that it is serving also as a ‘float coat’ albeit I am not looking to get an especially smooth finish. The depth is probably a minimum of 10mm, although it gets much thicker in places given the uneveness of the straw and holes or depressions in the first coat.
Most important learning is the degree to which the first coat will rapidly draw the moisture out of the second coat, making it quite difficult to work – hence the need to ensure that you give the first coat a good drink before putting the second on. It is amazing how much moisture the first coat will suck-up.
I am not using the gun, just loading the hawk up with a good load of render and then pushing this on and up using a trowel. I then quickly work this to get a reasonably smooth and even coating and finish off by misting it and going over the surface with a sponge trowel. This helps cover the trowel marks and produces a relatively even but rough textured finish – which has about the right amount of ‘hand-madeness’ I am looking for.
This doesn’t require a huge amount of skill and can be done quite quickly – not quite as fast as I could go with the gun, but not far off. I reckon I can get through about 7 square metres per day. This is a relief, since I was worried that my attempt at a finish coat would either require a huge amount of time and/or end up looking like a bodge-job. This isn’t to say that plastering could be considered easy, but it is not too hard if you are happy with a what might be called a rustic finish.
I know that time is ticking on and I am now well out of what traditional lime people would consider to be the rendering season. This might be the case if I was using sand, but with the chalk and fibre mix I am confident that I have something that will be resistant to all but the hardest of frosts and that if it looks like a cold snap is on the way, putting hessian up should provide enough protection. There has certainly been no evidence of any sort of cracking developing on either the first or second coats thus far.
The only other job this week has been to order the window cills, since as you can see from the pictures, I haven’t yet finished-off the area around the windows and am waiting to get the cills in before doing this. I have set the windows quite deeply, which means the cills need to be 220mm deep – i.e. much wider than any standard widths, so I have had to order the wood (oak) from a local sawmill. I am going to use just a flat plank and set it into the frame with a downward angle, resting on top of the straw/render rather than going to the expense of having a custom-made cill, with flat base and angled top since I figured this is only necessary if you are fitting your cill onto a flat surface such a a course of bricks.
Today was the first day of rendering. I must confess I was a little nervous. Things got off to a bad start when I found that the connection on my render gun didn’t fit properly on to the hose from the compressor. However, a lead from a local builder’s merchant unearthed a relatively local company called Airpower East, who specalise in all things air powered. Here a very helpful man searched through their extensive stock to find just the thingy required. It was amazing how many types of connection there are in the air world – so I would advise you to check your connections, especially if, as I did, you end up getting a render gun off the internet and are thus not sure exactly where it comes from.
The area I was planning to start on was the gable end at the side of the house that is least visible – based on the theory that it is best to practice on the parts that people are less likely to see.
Once underway I discovered that the key to making things run smoothly is to get the right balance between liquidity of the render mix and the amount of air you put through the gun. If the mixture is too runny and the air pressure too much, it just appears to blow through the mix without catching it, or else it picks-up the wet material and leaves behind the fibres, which then clog up the gun. A mix that has 10.5 litres of water per bag and an air pressure of 3psi seemed to be the sweet spot. At one point I had a mix that was perhaps a bit too wet and this kept clogging the gun, so I decided to try putting it on by hand by simply throwing handfuls onto the straw. Ths actually worked quite well, which is good to know as a back-up, but was pretty slow.
The product itself, Limecote, is amazing stuff. No matter how wet and sloppy you make the mix it holds its shape without any slumping, because of the density of the fibres within it. It is also fantastically sticky – especially useful when you are throwing handfuls of it at the straw. There is no way a traditional sand-based render would behave like this, or be as easy to mix.
The trickiest part was getting a good seal at the edges where the straw meets the steel underside of the roof. I did this by getting a handful of straw and rubbing render into it and then forcing this ‘long-straw mix’ into the join. This seemed to work, but exactly how this join will work long-term, I don’t know. Will have to keep a watch out for cracks / gaps.
By the end of the day I got one coat onto the whole gable end (probably about 3 square metres), which seemed like reasonable progress, given the late start.
More pictures and video to follow.
The Straw Works plans I am using specify facing the foundation box beam and the roofplate ringbeam with 20mm woodfibre boards to both improve insulation and create a base for render. This always seemed like a minor detail that would be easy to to complete – but has turned out to be more of a headache.
First-off, the diagrams specify a 20mm board and Straw Works have told me that they use a Steico board. However, Steico don’t stock an exterior grade 20 mm board, only a 20mm interior one. Their smallest weather-proof board is a 40mm with a tongue and groove edge – which is significantly more expensive that their interior board. I can see why T&G is necessary for a thickish board which is going to be applied in sheets, in order to get a good join, but I am applying it in narrow strips, onto a solid timber base. There is also another issue in that the straw doesn’t have a consistent edge where it meets the beams. Just in the nature of the stuff and the way the walls were put together in some places the bales are flush with the edge and in others it can bulge 40mm or more over the top. I did therefore think that a 40mm board would help even this out but might also create too much of a bulge in places given the need to have a 20mm layer of render on top.
In the end I found a 20mm, exterior grade, square edged board called diffutherm supplied by Back to Earth. However, something else I failed to appreciate is that the board itself is very soft. I had imagined it would have the consistency of MDF and thus hold a screw or nail. There is no way fibre board could do this, so you therefore have to use special flexible plastic washers (which Back to Earth supplied) and also stainless steel screws, since the lime in the render will corrode anything else. These were surprisingly difficult to source.
As it turns out, I will not need as many of these as I first thought, because in working out how to create a render stop and also install a board which will protect the tyres and act as a skirt to throw rain as far away as possible fom the foundations, I came up with a scheme which involves using a chunky 150mm screw to secure a 50mm square timber two-thirds of the way down the woodfibre board, drilling securely through into beam underneath. I can then use this as a base to attach standard black treated weatherboard as a combined render stop / rain skirt (see first picture) but it will also serve as a way of securing the lower part of the woodfibre to the foundation beam.
Straw Works specify that this timber should be an (expensive) durable timber such as oak – but I figured that weatherboarding would be fine, in that it is clearly designed as an exterior timber and it would also be very easy to replace whenever it begins to deteriorate.
It is also worth noting that Chris Brookman at Back to Earth was insistent that I would need to apply an additional (quite expensive) coating to the fibre if it was to be able to take a lime render. However, I checked this out with Martin, my lime man, and he said that might well be the case with sand-based render, but that his mix would pretty much stick to anything and that he had already applied it to diffutherm with no problems. In Chris’s words “I can’t say you haven’t been warned!” which is fair enough, but the Limecote render I am going to be using has essentially been designed as a material that breaks most of the rules of lime rendering – and this is just something I am going to have to take a risk with.