Roof Space

1. Schoolhouse Floor Plan

Table of Contents

Past renovations

Renovation Plan

There are no plans to turn the space into an attic.


The corrugated iron roof is quite old and uses slotted screws as anchors and has been repainted at least once, making the screws very difficult to remove. Fortunately, as I found, previous work on the building, e.g. electrical wiring and ceiling insulation, required access and two sheets of iron were removed on that occasion. Inside, there are no trusses, only rafters. It makes it very easy to move around in the space. Initially I assumed the rafters were joined to the joists but in image 2 below that is clearly not the case. Now I'm assuming that the top plate is anchored to the wall and that the rafters and  joists are nailed to the top plate - to be confirmed but I see no evidence of a problem, neither walls bowing nor ridge line sagging.

The opening gives access to the top of an internal  stone wall, wide enough to easily move along without treading on any joists, although later I'll need to. There are some ends of laths partially protruding over the top of the wall that I was also careful not to step on. There is wiring of different vintages around, some in conduit and some not.

There is plenty of evidence of bird, bat and rodent activity so a cleanup is part of the plan.

2. Looking at the top of an internal wall inside the roof space.

Ventilation ducting

The most curious thing in the roof space is the remnants of a passive ducting system that vented what are now the two bedrooms. There are register boxes above the centre of each ceiling connecting eventually to a single duct that rises vertically through the roof to outside. I don't know what this system was intended to do exactly but there were vents in the ceiling originally, usually installed to remove lamp oil smoke and fumes. The ducting isn't original but appears to have been installed in the early 1930's, and electricity wasn't connected to the property until 1952 or a little later so I assume it was for fumes as before and there was some need to duct it to the outside directly rather than into the roof space.

That need was sufficient to convince the Education Department to spend the money. Could it have been a government or council regulation change to improve habitability? Hence only the residential part of the building required it and the classroom left as built to vent into the roof space? The classroom and residence roof spaces are all the one space, so venting the residence also isolated it from the schoolroom. It may have been a noise issue but more likely a health issue. As I expand on these comments the world is in the midst of dealing with the COVID-19 pandemic and there are the comparisons to the 1918 flu pandemic. It's entirely possible that this ventilation was installed in response to that.

The system serves no function now except to vent the roof space itself and so the bulk of it will be removed.

3. The curious vent on the roof.

4. Directly below the curious roof vent. This is how I found it, with a section disconnected,  providing some roof space ventilation.

5. The other end of the horizontal section of duct. The register box would have been rotated 90 degrees clockwise from where it is here, in alignment with the other one in image 4.

There were vents in the ceiling already for oil lamp smoke and fumes but this ducting was not part of the original 1879 build. With the arrival of electric light circa 1952 the vents were covered and the ducting became redundant but left in place.

It is a passive ventilation system and I assume installed primarily to remove combustion products from oil lamps and possibly also to reduce condensation. When the open fireplaces were in use it may be that these vents reversed function and provided air from above to reduce drafts through windows and doors. There are no dampers to closeoff the ducts in winter.

By 1925 Crookwell businesses were installing electricity generators and equipment and locally generated electricity was being supplied to subscribers in the town for light and power. There's no evidence that there was any electricity generation at the schoolhouse. Once mains electricity was connected throughout the Cotta Walla region in 1952 or shortly thereafter and electric light was installed, oil lamps and the ceiling vents were no longer required. That alone must have made the rooms considerably warmer during the colder months.

Removing the major components was easy as they were all connected by friction fit only. Just a matter of sliding the joins to separate them. They came out through the same opening that they went in by, the sawn rafter still there after 90 years or so according to my estimation. I have left the last piece (image 3) protruding through the roof, which will act as it has for decades now to ventilate the roof space.

6. Getting the ventilation components inside required cutting a rafter 80 to 90 years ago.

7. They came out through the same opening used to install them.

8. The actual vent grill in the ceiling from above, now covered on the underside by the ceiling rose.


Fibreglass insulation batts have been laid between the ceiling joists throughout the roof space. There are pink batts under the yellow batts and the colours indicate different manufacturors, so my guess is that additional insulation was added later, but not much later. Combined the two layers of batts would have had a R value of from 2.6 to 3.3.

There are some sheets of what look like solid core insulation stacked on the ceiling over the classroom (image 9). I suspect I will find that this type of insulation has been placed between the vaulted ceiling panelling and the roofing iron, were it wasn't possible to use the batts.

9. Insulation extends through into the vaulted ceiling over what was the classroom (now the lounge).

An access panel in this sloping side of the vaulted ceiling would be useful, with a landing on top of the stone wall and a bridged walkway across the bedroom ceilings.

Ceiling and electrical repairs

I'm going to remove all the insulation above the two rooms that have lath and plaster ceilings and get the workshop vacuum in to clean up. This will let me do any repairs needed and then I'll probably lay new R5.0 insulation.

The electrical wiring is all post 1952, which is quite late but allowed the industry time to establish solid wiring standards and regulations by the time this building was connected. The only real issues are the risks posed by rodents that could knaw on the cables.

Insulation removal

Removing and bagging the batts was dusty work and who knows what microlife is mixed up in it so I'm glad I wore a good dual cannister respirator mask. The installer left the paper packaging with the product description on it as "extra" insulation. The pink batts are R1.3 and the yellow batts on top are a little thicker so I'm guessing they were R2.0 (R3.3 combined).

I used the hardwood plank at far left of image 10 to support me, for two reasons: I'm 115kg and I don't trust those 140 year old joists; and, I don't want to flex the ceiling too much and cause damage.  I'm standing on top of the dividing wall to take those photos.

I found three rat skeletons and several nests and burrows in the insulation.

There are no obvious ceiling problems from this viewpoint but all I can say really is that there are laths and plaster keys (A.K.A. nibs, lugs, snots) but is there a ceiling hanging from them? I know bedroom two is ok except for a small area that is sagging slightly. As for bedroom one I can't be sure without looking at the underside of the ceiling, which means cutting some holes in the lower drywall ceiling in that room.

Good to see that the lighting wiring, which appears to be an active loop at the light installation, is mostly in conduit but there are sections of power wiring that isn't.

The next job is to vacuum.

10. Batts removed over bedroom one

11. Batts removed over bedroom two

The 1"x1" strips of timber running across the joists at left of centre in image 10 and top right of image 11were used to carry wiring clipped to the strip before conduit was used. It suggests the building has been rewired at least once. They are now superfluous and can be removed.

The 3"x1" strips were originally one continuous length (known as a "binder") installed across the middle of each ceiling evidenced by the nail holes in the joists and I assume were cut to make way for the register boxes over the vents. As used here the binder's function would have been to hold the joists in place laterally until the ceiling was completed. The modern term for these in Australia would be "bottom chord tie" while "binder" has acquired a more structural definition I believe.


My workshop vac has survived a lot of lime render dust.

I'm not concerned about lead particles because traffic on the road outside is very low and I'll be wearing the respirator again.

Spent about four hours to complete about one third of the ceiling over the two bedrooms, or about 10 sq m.

The debri is a strange concoction. Obviously there is dust but a kind of felty mat has built up that requires it to be physically disturbed before it can be sucked up. There is also a lot of vegetative matter and other small building debri like wood chips, nails and wire offcuts that tend to block the vacuum. These are picked up by hand before vacuuming the area.

My vacuum has three filters, the first pleated paper filter collects everything but some stuff gets around this filter and if it doesn't get right through the machine to the outlet it lands in either the thin foam sleeve or the small foam plug that prevents anything too big going through the mechanics and that I discovered today was very clogged. Cleaning the plug filter made progress another 50 percent quicker. I'm having to clean the primary filter every square metre or two and it was already looking quite battered. What I need is a cyclonic extractor.

I'm having to do one lath strip at a time between the joists and need to be careful not to break any of the plaster keys with the nozzle of the cleaner.

I can see that the laths are split timber strips which, just to note it here, are  stronger than sawn strips.

Another four hour session and one whole room ceiling done and a quarter of the other ceiling. Five more hours will finish both ceilings so luckily I find it kind of therapeutic, like gardening. Thirteen hours in all including access and setup time  on each of three days to produce thirty square metres of gently and carefully vacuumed ceiling.

Vacuuming completed. I estimate that nearly 40kg of dust and debri was removed.

12. One hundred and fourty years of dust and three minutes of vacuuming.

13. Bedroom one - vacuumed

14. Bedroom two - vacuumed

Summary of repairs

Now that the space has been cleared and I can see what's what, the list of repairs includes:

In bedroom two the ceiling has a localised bulge of several centimetres (3cm to 4cm), only really seen by running a straight edge across. This corresponds to the joist second from top in image 14, which has a localised downward bow in it, roughly between the 3"x1" strip at centre and the plank at right. There is a smaller area in this region where the plaster appears to move slightly when pressed. The joist has been in this condition for a very long time, perhaps always, so attempting to straighten it risks damaging the plaster that has flexed over that time or was applied originally to the existing contour.

There is a similar bulge in the bedroom one ceiling from the same cause, i.e. a joist that has a bow it. Likewise, attempting to straighten it risks greater damage and it is not clear that it occurred after plastering but this view may change once the original ceiling can be seen from the underside.

There is no reason to think these deformaties or their effects will worsen so the repairs will focus on reinforcing the connection between the lath and plaster from above and then patching the hairline cracks from below. One obvious area is around the vent where the keys are nearly all broken off. Over bedroom two there is already a copper wire tie around the joist to support the weight of the ceiling rose and pendant light below.

Ceiling plaster keys reinforcement

A test area will validate the method, which will be to soak the [lime] plaster keys and lath in areas of suspected broken or weakened keys with a diluted PVA glue solution, 3 parts water to 1 part PVA glue, a ratio commonly used on masonry walls before plastering (not to stick the plaster to the wall but to reduce the suction of the wall materials). Judging by the frothing that forms when mixing vigorously the glue appears to have some surfactants in it already or the PVA itself is a surfactant (PVA is a hydrophilic polymer), which is good because that will also aid penetration into the plaster and is what makes it a good wood and ceramics glue. I didn't see any signs that an additional wetting agent was needed but did research it and would probably try something like "GOLDEN Artist Colors Acrylic Flow Release" before trying Floetrol. The former product is more likely to be found in artist supplies stores than hardware stores.

The diluted PVA should bind plaster that has, or started to, become fractured, particularly in the gaps between the laths where the keys tend to break. In the end, as a preventative measure, I decided to apply the treatment to the whole of the ceilings. From the test area the total volume required was estimated and then mixed as a single lot in a bucket. The job took a total of 7 litres of PVA glue (28 litres of the diluted emulsion) over approximately 30 square metres. I did not need the cross-linked PVA product at three times the price. Instead I paid AU$26 per 4 litre container.

A much larger consolidation of a church ceiling used the same method and similar emulsion, so I have confidence in the result. As the restorers in that case observe, "one of the interesting ironies of the ... treatment method is that the worse the condition of the plaster, the better the treatment works ... because badly deteriorated plaster is more porous and receptive to the application of acrylic resins".

They used an airless paint spray machine to apply their solution, and although I have one I didn't think of using it. It would have been too much setting up for my small job anyway and too awkward to operate in the space. Instead I used a small 2 litre common plastic garden spray bottle with pressure pump that I could hold and operate with one hand so I could use the other hand to crawl or squat and steady myself. The small bottle also allowed me to better judge coverage at approximately 2 square metres per fill.

There is some evidence that this method has worked very well.

15. Water/PVA glue mix 3:1 applied as a heavy spray.

Span at bottom was done first and it can be seen that the glue has largely soaked in. This is a cold day around 13 deg C and very little evaporation.

Areas of sagging ceiling

There are no areas of major concern from below and only one area so far identified from above where there is a gap of 2mm between the lath and the joist. Actually not the joist itself but a short batten fixed to the side of the joist used to support the ends of some continuation laths. The batten supports are also loose and need fixing before glue is squeezed into the gap and the ceiling gently pushed up from below. This is the ceiling above bedroom one so will need to wait until the lower circa 1983 gyproc ceiling is removed.

Where the plaster itself is sagging, a water based construction adhesive will be applied via holes drilled through the lath from above into the gap between the laths and the plaster base coat, then gentle pressure applied to the area from below until set (24 hours).

16. A batten used to secure the ends of continuation laths.

The batten allows the end of a short lath to be nailed to the joist and the end of the continuation lath to the batten, I assume reducing the risk of the lath splitting because of a lack of "wood" between the nail and the end of the lath.

Installation of the new ceiling insulation

After the parlour ceiling light wiring was replaced and enclosed, new metric R3.0 (R17 in US terms) batts were laid.

There was no balancing on joists during the process for fear of flexing and cracking the newly repaired and painted ceilings below. Two hardwood 200x50x3000mm planks were laid across about 1/4 in from each end of the joists then a 3m aluminium plank laid between them to spread my weight. It was a little more effort to set up but a lot more comfortable to work.

Higher R rating batts (up to 220mm thick R5.0) were impractical with the existing wiring and other structure running across the joists. An additional layer may be installed later but would cover the joists and wiring completely so, ideally, a suspended walkway would be installed first. That would take the R rating up to metric 6.0, or R34 in US terms.

There is no insulation that is rodent proof, however, polyester heat-fused batts were used here because the material is claimed to be more resistant, but this install is experimental as this stage. The old fibreglass batts (two layers) were mostly still intact but filthy with dust and vegetation debris after twenty to thirty years. The polyester batts were certainly much more comfortable to work with.

17. New polyester metric R3.0 batts