Earth Notes: On Super-Insulating Our Smallest Bedroom and a Half (2011)

We'd already insulated better above it by topping up the loft insulation. Below it we had insulated the living room ceiling. We also reduced air infiltration from outside with foam.

Doing just this one small wall should not require building control (with October 2011 part L reg applying). Especially as our target was a U-value of 0.3W/m^2K or lower. Though I did exchange emails with building control (BC) at the council back in July to make sure.

Magnaline Superslim

This time, my aim is in part to provide a case study to help others plan (or plan to avoid, I suppose) aerogel-based IWI/drylining. So I went with a different brand of board, Magnaline Superslim magnesium board from Enviroform Solutions.

(Magnesium board aka magnesium oxide wallboard aka MgO board, used in place of traditional plasterboard AKA gypsum drywall.)

I wanted to reduce the chance of thermal bridging by the partition wall between the boy's room and the girl's (larger) bedroom next to it to the outside. So a slot was cut in the partition wall between the rooms next to the exterior wall. The aerogel was continued though that along the west wall to the edge of the north wall in the girl's bedroom.

It requires only 50% more drylining material to achieve this.

In that way both those small rooms had their west-facing exterior walls drylined at the same time. At some point in the future we may dryline that (larger) north wall to vastly improve overall thermal performance of the girl's room. (We did!)

This time, as our bedrooms are heated less of the time, and to a lower temperature, I went for 30mm of aerogel rather than the 40mm in the living room. We did our calculations based on the assumed external wall construction and its implied U-value of ~0.8W/m^2K. That should bring us well under the regs' 0.3 U-value.

Wall Buildup

On exploration we discovered that there was less mineral fibre than we had believed, 25mm rather than 50mm. That results in a U-value more like 1W/m^2K from this construction:

Removal of the existing plasterboard (to save space in a very small room) gets the wall to pretty-much spot on 0.3W/m^2K, >3x better:

The board has a foil VCL in front of the aerogel (behind the magnesium board). All gaps were foamed, then the wall primed with Gyproc Drywall Sealer to form an inner VCL. All with the aim of preventing moisture migration through the board and condensing in the woodwork with the risk of rot.

The builders reported the board to be easier to cut and drill (less snagging) than the Spacetherm. However, these boards were more difficult to fix/level because the reduced 'give' of the magnesium board caused screws to pop and not lie flush.

(I believe that the aerogel in each case is the same Aspen Spaceloft blanket. Though presumably slightly reformulated as this stuff was blue/grey whereas the Spacetherm was white.)

We also avoided penetrations to the outer wall by removing a socket in the larger bedroom, and surface-mounting power sockets in the both rooms. We moved the radiators to (opposite sides of) the internal partition wall to avoid pushing heat out through the external wall quite so fast.

Curtain rails were mounted on wooden battens to minimise point loads on the magnesium board.

Magnaline insulated window reveals (containing 10mm of aerogel) were fitted in both bedrooms, with a pine architrave/reveal fitted over that.

There was quite a lot of air infiltration visible around the windows when work was in progress. Non-expanding squirty foam was applied to fix.

Monitoring

Using iButton miniature temperature monitors to check thermal performance revealed a severe air-infiltration problem under the floor in the room before the work had even finished. That lined up with evidence of infiltration above the living room ceiling (ie the same void) in thermal imaging. Foam was applied to reduce that leakage.

As of 2014 the house is being monitored in real-time. For example a sample from overnight 2014-11-13 showed that this bedroom's temperature (2b green line in the graph) stayed steady overnight. It was at equilibrium with one occupant (~100W heat input) and a temperature differential of ~6.5°C from outside. (Ignoring heat flows to the rest of the house.)