Rendering Magnesium Oxide Board

The laundry and second bathroom had been left unrendered as these rooms had tiled floors and did not need to be done for the polished concrete floor to be finished. Further delay was caused as I had not finished the mosaic tiling for the bathroom. With Christmas coming up and more people staying in the house, it was time to finish the second bathroom.

I had been unhappy with the finish of the AHMC render mix on the MgO board walls and the renderers had found it very hard to get the render to stick to the MgO walls. I raised my render on MgO issues with some other hemp builders, sought their assistance and received several suggestions. I had also seen some straight sand lime renders used on strawbale houses and with all this I undertook my own rendering experiments.

Freshly done render tests on MgO Board

I tried three different mixes applied using two different methods. The top row from left to right are (1) standard AHMC hemp render mix, (2) half AHMC hemp render mix half lime/sand mix and (3) straight lime sand. The lime/sand mix replaced the same volume of AHMC render with straight builders lime. The bottom row uses the same mixes but with the addition of a coat of Bondcrete on the MgO board left to go tacky. The Bondcrete did not work at all and in fact it was harder to get the render to adhere to this mix than to the untreated MgO board.

Dry render tests on MgO board

(1) standard AHMC hemp render mix

(2) half AHMC hemp render mix, half sand/lime

(3) sand/lime

 I had a bit of mix (2) and did some extra sections at either end of my test MgO board. I have some disclaimers to this test as firstly I am not a good renderer, secondly the test was carried out on a hot day and thirdly the number (3) mix had too much lime to sand and should have had more sand, the replacement of AHMC mix with straight lime should have been done by weight rather than volume as the AHMC mix had hemp fibres in to which are very light but bulky.

Mix number (1), the standard AHMC mix, was still difficult to adhere to the MgO Board and as on the MgO board the mix can not press into the wall like it can with a hemp wall, the hemp fibres came to the surface and stuck out too much for my liking. Mix number (2), the half AHMC half sand/lime mix was easier to spread and adhere to the MgO Board, it also produced a smoother surface. Mix number (3) the sand/lime, mix cracked badly when  applied at the same thickness as the other mixes and was hard to apply in a single 8-10mm coat. I hoped to be able to get away with using a single coat to save labour and hence cost.

John, my first renderer, came around and looked at the tests. He agreed that the second mix looked best but we agreed to use a little less lime in the sand/lime mix. The mix we used was two bags of 22.35kg of river sand (this had been measured out earlier for the render and we had some measured out bags left over), one 9kg bag of AHMC hemp render mix, half a 9 litre bucket of builders lime (so about 4 1/2 litres) and 160g of yellow oxide (this was double the amount used for a single AHMC mix, since in effect we were making a double mix).

The new double size mix of render was mixed up in our pan mixer (mortar mixer) that we had used to make the wall mixes. Previously we had used a standard cement mixer for the render as out pan mixer was out on loan to another hemp builder. The pan mixer was a bit slower than a cement mixer but did a good job mixing the render and was much less dusty.

The amended render mix adhered to the MgO board much better and easier to move around on the wall. The MgO was still not the greatest substrate to render on but it was now much more achievable. A smoother more consistent finish was also able to be achieved. The only issue was that the walls were a little yellower than expected. I really should have tested the actual mix we used first rather than just relying on my test. I assumed it would be ok as my test used a proportional amount of oxide, but as we now used less lime the colour became more intense. The good thing is that the family liked the more intense colour and it is a nice bright sunny yellow in these small south facing rooms.

Amended render on MgO Board

Rendering the laundry

The bathroom also had one hemp wall. The amended render went on this wall really well. When John started rendering this wall I wondered why he finished so quickly, it was because even with the amended mix it was still much easier to to render a hemp wall than a MgO wall.

The last walls in the bathroom were rendered between Christmas and New Year. It was great to be able to hang the bathroom mirror. When we were building the walls we put an extra nogging in the frame right where the mirror had to be hung and thankfully we remembered where it had to go and the screws hit the nogging.

Bathroom mirror on wall

Freshly rendered bathroom wall

The only disappointing news was that we were one bag of render short and one wall in the laundry is still unrendered. This will be fixed up in  the new year. While disappointing it was better to leave this wall than try and do a too thin coat of render and have it crack or not work well.

How to Hemp Lime Render

With well over half of the internal walls rendered, John our renderer has mastered the application of the Australian Hemp Masonry Company hemp lime render. We have put together a little video on the rendering process.

The first room to be completely rendered was the bathroom. In the bathroom the render produced a very nice finish that ended flush with the tiles, each being about 10mm thick. The main kitchen living area was the second room to be completed. Two of the bedrooms are not far off completion with only the second coat on the Magnesium Oxide board to go. Even the third bedroom now has one wall rendered. All the hemp walls have been rendered in a single 10mm coat. The MgO board walls have been rendered in two approximately 5mm coats.

Last wall in the main room freshly rendered

Main room rendered

First wall started in the last room to be rendered

Progress Made on Hemp Lime Rendering

We bought our pre mixed bags of render at the same time we bought the binder for the house. The lime render came with just a list ingredients to add and instructions to wet down the wall before applying the render. I had never used the AHMC render before I bought a pallet load of it for my house. I had done a brief bit of hemp lime rendering with Steve Allin at his workshop in St Ives in January 2012, but they used a different binder, very large hemp fibre and put on a layer almost 1 inch thick. This process was so different to that done at the St Ives workshop that it was of no use when it came to the hemp lime render we were using. My experiments to get the right colour and sand only gave me experience in doing 300mm square patches and even then the render sometimes went on really well and other times was really difficult.

Thankfully we had John our renderer to help us out, so these issues of getting the right mix and consistency were placed in his experienced hands. The first test wall John did, inside the wardrobe just in case things did not go well, was slow going but worked out fine. The next wall was faster but still quite slow. He quickly learnt that the hemp lime render does not spread easily across the wall, whether it is the hemp fines or some other reason, it is best to apply the render as close as possible to the location you want it and as close as possible to the finished thickness, then just use the float to smooth and even it out. The steel float gives a nice smooth finish, but the pieces of hemp and some of the large pieces of river sand ensure that the render maintains some texture.

The order of mixing using a standard cement mixer was ¼ bucket of water, ¼ of the 22.35kg bucket of river sand, ½ a bag of AHMC render mix, 80g of yellow oxide, the rest of the water and the rest of the sand and the rest of render mix. Mixing was a bit dusty but the holding of an empty bag over the mouth of the mixer after the render mix was added kept a little bit of the dust down. Later we changed this to using a bit less water and then adding a spray of water from the hose after the render mix went in and this kept the dust down.

The first part of the third wall went well then a slightly stiffer mix slowed John down substantially. He was not able to achieve the same finish as the first half of the wall, so he had to damp down the stiffer render, use a sponge float to work this water into the render then steel float it all again. Lesson learnt that there needs to be consistency in the amount of water in the mixes and the slightly sloppier mixes go on better. Achieving consistency in the wetness of the mixes is not as simple as using the same amount of water, as some buckets of river sand contain more or less water this needs to be adjusted for. So, like cooking it is about getting the look and feel of the mix right.

Front pavilion one day after rendering

Left wall fully dry and right wall 12 days after rendering

3 weeks after rendering

Wall freshly rendered

12 days after rendering

3 weeks after rendering

The weekend John worked on the third part of wall, my mate Tony was on site and having been to two hemp building workshops himself, he was keen to give hemp rendering a go. Tony and I got to work on one of the small MgO Board walls on the inside of the wardrobe. I used the stiffer mix and it was very slow going, in the confined space of the upper wardrobe it was awkward to trowel the render on, so I rubbed handfuls on the wall pushing them on until they were the right thickness then giving it a trowelled  finish. Given that John was at that time rectifying the effect of the stiffer mix Tony experimented with using a much wetter mix. The much wetter mix went on faster and could be put on with a trowel, however in parts it had difficult adhering to the wall. Going back several hours later to fix a patch where the wetter render had slumped a bit, a whole patch fell out. It appears that you need the “just right” mix, not too sloppy and not too stiff.  

The skinny wall done and patched by the non professionals

The following day John only had a couple of hours to spare and finished one of the walls with lots of windows. I used the “just right” render left over from John to patch the hole in the render on the side of the wardrobe. I had sloped the edges of the patch that had fallen out so there was less of a line where the join would be. Wetting down the edges of the patch and working the new render into the old, by making circles on the wall with my rubber gloved hand, the patch was filled and while it was not prefect  it was acceptable. John had also left a small join next to one of the doors and when he finished it the following day had to feather in the new mix.

With still more render mix left over I worked on the upper inside section of the wardrobe. A ridiculous section to render as no none will ever see it, but a good place to practise. I was painfully slow at rendering and it was very hard work on my arms, but the finished render was acceptable and was sticking to the MgO Board walls. Given the snail like speed of my progress I am glad to have outsourced the rendering.

We have applied self adhesive render mesh diagonally across the corners of the windows and along the joints between the hemp walls and the MgO walls as well as along the joins in the MgO Boards. I am using a 5mm mesh that comes in 200mm wide rolls from Foamex. The render mesh is to try to restrain any movement within and between walls to prevent cracking. Despite using the render mesh we did get some hairline cracks at the corners of the first window we did. Later window corners had either no cracks or only very small cracks.

Only a very fine crack at the corner of the window

No crack on the other corner of the window

We used 200mm x 300mm strips of render mesh across the corners of the windows, however had some adhesion problems and these sections of render had a tendency to pull away from the wall while they were still wet. These patches were identified and fixed while the render was still wet. The cause of the problem was hypothesised to be the hemp fibres in the render preventing the render from pushing through the 5mm grid in the render mesh and coming into contact with the wall. For further corners we tried halving the width of the render mesh to 100mm x 300mm across the corners of the windows and doors. This resolved the adhesion problem and when fully dry after about 4 weeks we will see if there is any cracking. The adhesion problem was not such a big issue at the corners where 200mm wide render mesh was run along the joints between the hemp and MgO walls, but here only 100mm wide on each wall was being rendered at each time. 

I had originally thought that we would need render mesh to cover the whole of the MgO Boards so that the render would adhere to it so I bought from Foamex (manufacturers of polystyrene walling and insulation systems) two rolls of 200mm wide mesh and two rolls of 1200mm wide mesh. I had believed that both widths were self adhesive, I checked the small rolls when i purchased them but the larger rolls were wrapped up and I just assumed they were the same. But, when I got them to site I found that the 1200mm rolls were not self adhesive. Going back and checking my original quote I found that the mistake was mine and the 200mm rolls were listed as self adhesive and the 1200mm were not. I confessed to my mistake and e-mailed Foamex asking if I could return or exchange the 1200mm rolls. As they were unopened and unused this was not a problem. I went back to their warehouse where I had originally picked up the mesh from and had no problems at all returning the 1200mm rolls, getting three more 200mm rolls and getting a refund for the difference. The fabulous service I received from Foamex could not be in more contrast to the problems I have had with the plywood. 

While I was off site John continued rendering, as rendering is his second job it had to fit in around work and family commitments. John did three walls of western bedroom in the back pavilion and a section of the external wall, before he was struck down by illness and forced to take a break. It was interesting to watch the process of the internal walls drying. For the first three days after they are done it is hard to see much visual evidence of drying, although you can feel the render firming up. After about a week the walls look unpleasantly patchy as different parts of the wall dry unevenly. Two weeks later the wall looks much more even and after three or four weeks the render is all dry and stops changing.

Back pavilion a few days after rendering

About a week after rendering

Right wall done later than left

Rendering our first external wall was good and bad. Without the ceiling and floor detail to worry about the rendering went quickly and we chose to have a slightly rougher texture on the outside. Problems arose when a chunk of hemp wall next to the window fell out. I had taken an angle grinder with a masonry blade on it to the walls and taken off the sharp corners around the corners and around the windows, so that I could get a nice curve around the window frames. However, I believe that this cutting off of the corner may have cracked a weak section and we will not have to try and patch the section with render. Next time we will try rendering around the window without cutting the corners off the hemp walls.

Detailing around the window

Rendering the external wall

Where a piece of the wall fell out

Window detail

External wall a week after rendering

Second hemp render test and Plywood Problems

Three weeks after our first render test we got to see the dry colour - grey. Not a successful test. Anticipating, from the colour when wet, that it may not be the sandstone colour we were looking for I sourced some dark yellow river sand from Thomas Landscaping at Lawson and they gave me a small plastic bag of it for free to test out if we liked it. So this became render test two.

Dry Sydney sand render test on left

Dry hemp lime render mix before sand added

 The Australian Hemp Masonry Company, who made the render mix, recommended using river sand but I first tried using Sydney sand (Beach sand) because local river sand in the Shoalhaven is a dark grey, not the colour we were after. I really liked the colour of the new dark yellow river sand, but when mixed with the render mix it got the grey/green look of the previous mix, albeit this time a bit yellower.

Yellow river sand render test showing colour of sand used

I am yet to see the second render test dried, but am ready to give up on the hope of being able to colour the render with sand alone and will now try some oxides. AHMC advise that the hemp wall need to be wetted before the application of the render, this was a bit tricky for our first render test, but for our second render test I used a garden pressure sprayer, of the type usually used for fertiliser. It worked well giving a fine spray of water that meant it was easy to control the area that got wet and how much water was applied.   

Pressure sprayer in front of render test

 Meanwhile work continued on the plywood ceiling. Despite one wall of the front bedroom not being square, we managed to do quite a good job on the ceiling.

Front bedroom ceiling finished

Further work was then stopped as the BB Hoop Pine ply had some horrible blue marks on it. Both on the back and face surface. We looked through the remainder of the pile and managed to find two sheets with which to finish the front bedroom. The rest of the sheets had blue marks that looked like mould, to use it on the ceiling would look like the roof had leaked and the ceiling was going mouldy, not what I paid good money for B grade ply for. 26 sheets had unacceptable blue marks on them.

Blue marks on ply

Stack of blue marked ply

A call was made to the supplier Huskisson Ply late on Friday, but the boss was not in, so no response. Several phone calls later, I finally caught the boss on the following Thursday. I e-mailed off pics of the blue stained ply and they will see what they can do. Between trying to get through to Huskisson Ply I also phoned the ply manufacturer Austral Ply to find out about the blue marks. It seems that the Australian Standard says that they are acceptable for a B grade ply - but who reads the standard! I thought by paying the extra for a B grade that the face would be good enough to use for the ceiling. None of the pics or info on the Austral Ply website mentioned blue marks. Apparently they can come about by the logs being too wet before they are made into ply and starting to go mouldy, while i was told that the ply production  process kills the mould - this does not help the fact that if used it would look like my ceiling was mouldy, not acceptable for a new house. Austral were very helpful and suggested I get the place I purchased it from to contact their supplier and see if they could exchange the problem sheets of ply.  

Meanwhile with no more usable ply sheets, we worked on the blocking that goes behind the express joints in the main room. Whether it was experience or working in a larger room but the work seemed to be getting easier. I have also accepted that nothing is ever perfectly straight in building and that you need to measure the exact location of where each piece goes and cut it accordingly. When I first started the blocking I had some problems skew nailing it with the framing nailer, particularly pushing the nailer in far enough to fire and getting the right angle. Now I can do this with ease, even using my right or left hand.

Blocking for express joints in ceiling

If necessary we can also move on to starting the internal stud walls since the Magnesium Oxide Board finally arrived. We are mostly working on site on weekends so originally booked in for the MgO board to be delivered on a weekday we would be on site. That day arrived and when it had not arrived and we had not heard about the delivery by mid afternoon I made some phone calls and found out that my instructions that the delivery had to be made on a particular date were not passed on to the transport company.

There were many subsequent phone calls as the delivery company could not deliver on weekends and could not deliver without us on sit  as we had to unload it, so because of other commitments we had that meant we could not be on site on a weekday, our delivery of MgO Boards arrived four weeks after we had originally booked for its delivery.

The MgO sheets are 8mm and did not look too heavy, but looks were deceiving, I was only able to carry one sheet at a time together with another person. Further because the truck was side loading it could not go down the driveway, otherwise there would have been nowhere for it to unload. We started carrying sheets down the driveway one by one. This would have taken a long time as we have a battle axe block. Things were sped along when the delivery boys started using a trolley loaded with four sheets at a time to wheel the sheets to the end of the driveway, from where Ben and one of the delivery boys were able to carry them two at a time into the house.

MgO Boards stacked in the house

Why did we use MgO boards - they are breathable, not damaged by water, dimensionally stable, do not harbour mould or mildew, are not eaten by termites, insects or vermin, are non toxic, and I had seen them used on Roger Bodley's hemp house in Tasmania, where they seemed to work well rendered as internal walls, which it the application I will be using them for. 

First Hemp Render Test

We did a test patch of render so that it will have plenty of time to dry, before we have to choose the sand we use for the render and whether we need to add any coloured oxide. For our first test we used the Australian Hemp Masonry Company bagged lime based render mix, we chose this for the ease of using a bagged product which would be consistent through each mix and because we like the fine hemp fibres in the render, so you can still see that it is a hemp house. To this we added  water and sand. Made up 1/20th of a 9kg bag, so ended up adding just over a kilo of sand. We used Sydney sand. AHMC recommended the use of river sand, however the locally available river sand in the Shoalhaven area is a dark grey. This was all right for the wall mix but is not what we want for the colour of our render. I was told by the landscaping and hardware store that the Sydney sand was what was used for rendering, that of course referred to cement rendering. The bulk Sydney sand I had seen was a light yellow, but when a bag of sand was purchased it was browner than expected.

Mixing the render in a bucket

Because we were mixing a small amount we just mixed it by hand in a bucket. We chose to do a test patch behind where the kitchen cupboards will go. The hemp fibre made the mix stiffer and less plastic than a cement render. Using a hawk and float it seemed to stick to the wall quite well, although pressing the render into the wall gave that arm muscles a good work out. The wall had to be damped down before the render went on. This was a bit tricky without a sprayer and we have subsequently bought form Aldi an 8L pressure sprayer, hopefully we can damp down the internal walls with this, so we do not have to use a hose inside the house.

I have read elsewhere hemp lime render being described as like "tuna and mayonnaise", the fibres in the render certainly made it different to work with than I had expected. With cement render, it is usually put on and left for about two hours before it is floated again. We did this with the hemp render, but two hours was probably too long. The stiffer mix was too hard to push out anything but the smallest lumps. At Steve Allan's workshop he had suggested rubbing the hemp plaster mix back with a rubber glove, when it was partially set to bring the hemp fibres to the surface. He was using a different render mix, that had large hemp fibres and no sand. We tried this out, but using the AHMC render mix this brought out too much of the fibre and I prefer just a floated finish.

Test patch of render

Render rubbed more at top of the patch than bottom

 Our rendering was a bit lumpy, but with some depth guides and a bit more practise I think we can get an acceptable finish. I would still like to get some lime rendering lessons if I can find someone willing and able to teach us. We will not be able to tell the colour until it is dry, but I have sourced some yellow river sand a bit further away, but if it gives us a better colour without having to add oxide it will be worth a couple of trips with the trailer. We will try this sand next time we are on site.

Meanwhile work on the ceiling continued. We returned the broken panel lifter, the problem was very obvious and they exchanged it for a new one. The paint job on the new panel lifter is a bit patchy but the mechanics seem fine. We put ti to use finishing the second bedroom in which the last ceiling panel had to fit into a recessed section over the door that had walls on three sides. This was tricky, but after cutting a little off two of the edges it fitted acceptably well.

Ceiling in second bedroom

Work then progressed to the last of the bedroom ceilings. Following the line of the eave rafters a ceiling board would have ended just short of the built in robe. The robe was also 10cm narrower than it should have been, another mistake by the framing company, but too difficult to fix now that the frame is all up. The solution was to add some extra blocking behind the ply ceiling boards so that the end board lines up with where the ceiling abuts the robe wall. The brick pattern was continued for the ceiling in the largest bedroom even though this meant that every second row had two small boards on each end. The alternative of one big straight line down the middle of the room would have looked odd and shown up any slight difference in the spacing of the boards.

Ceiling in largest bedroom