1 Contacts

For further information please contact:

2 Context

The home consisted of a weatherboard dwelling of ca. 1910 vintage with a 1980s extension to the rear. The block is oriented such that the narrow frontage faces north, reducing the opportunity to maximise passive solar gains. Moreover, the area is covered by a heritage overlay which precluded extensive alteration to the facade. This included retention of the two brick chimneys.

Figure 1: General views

3 Our ambitions

We’d lived in northern Europe for a number of years and become accustomed to buildings that have a sense of solidity and stable internal temperatures. Neither of these features seem common to Australian homes in our experience. Moving back to Australia we bought a similar weatherboard home in 2007 and retrofitted that home with double glazing, solar PV, additional ceiling insulation and pumped expanding foam insulation into the external wall cavities. This new home provided an opportunity to go much farther, as it required extensive renovation.

As well as aiming to meet our comfort expectations we were keen to minimise our energy consumption and move towards energy self-reliance as much as possible. To achieve this we needed to ensure our energy needs within the building, for which space heating is by far the most important, would be minimised.

4 The advantages of Passivhaus

There are numerous sustainable building philosophies, most of which seem to have a strong “feelgood” basis and much more limited technical rigour to support them. We liked the passivhaus approach because of its scientific basis:

While passivhaus is fairly common in Germany, and related standards such as Minergie-P are common in Switzerland, it’s fairly new to Australia. Indeed, when we started there were fewer than six certified passive houses in Australia and only one certified to the retrofit standard.

We emphasise that our home is NOT passivhaus certified. In particular, we feel just short of the airtightness criteria.

4.1 How is passivhaus different?

Australian building practice increasingly recognises the importance of bulk insulation as a means of maintaining internal building temperatures, and there is some understanding of the importance of double glazing. There is also some recognition of the role thermal mass (e.g. concrete slabs and reverse brick veneer) can play in storing and releasing heat within a home. However, there is almost no recognition of:

  • the importance of airtightness,
  • thermal bridging, or
  • verification of modelled performance.

The passivhaus approach is different insofar as it requires extreme levels of airtightness, a continuous insulation layer and verification that what has been designed has indeed been built. More information on the approach is available here and from the Australian Passive House Association.

5 Technical summary

Internal volume 495 m3
Airtightness
Air changes per hour @ 50 Pa (ACH50) 1.20
Mechanical ventilation and heat recovery (MVHR) Zehnder Comfoair 550
Insulation
Ceiling R8.0
Suspended floor R2.9
Rammed earth wall R4.0
Timber framed walls R4.2
Glass R1.4 (U0.7)

6 Insulation

Conventional glasswool insulation batts were used underfloor, in the ceiling and walls. Higher density batts were chosen as they (a) have additional stiffness such that they are less likely to sag over time, and (b) provide marginal additional sound absorbing properties. Extra care was used in placing the batts to ensure there were no gaps and that the batts were not unduly compressed. Recall that the performance of insulation is not due to the material itself but rather the capture of many small air pockets within the material. If a glasswool batt is compressed there will be far fewer air pockets and hence lower insulation values.

Figure 2: Typical wall insulation - this is the second layer inside airtightness membrane, studs are offset from outer wall

A conventional timber stud wall construction is 90 mm thick. This provides inadequate depth to achieve the levels of insulation we required. In the older part of the home we installed an additional 90 mm stud wall internally and placed the airtightness membrane between these walls. Each wall was filled with conventional R2.7/90 mm batts. This construction provides an R-value of around R4.2 (accounting for the presence of the timber framing). In the new extension the double-stud construction was also used.

By placing the airtightness membrane in the middle of the wall thickness we risk condensation forming on the inner membrane face in winter. While the modelling suggests this risk is minimal it would be better practice to place the membrane towards the inner (warm) side of the wall construction. However, one notional advantage of the method we have used is that the membrane is 100 mm behind the inner face of the plasterboard, ensuring we’re unlikely to accidentally pierce it in the future.

6.1 Thermal bridging

A thermal bridge occurs in a construction where there are multiple materials of differing thermal conductivity such that heat can readily escape (or enter) through the material with the highest conductivity. When building for efficiency this becomes increasingly important. Around 15% of a typical timber wall construction will be timber framing. Timber has a thermal conductivity around three times greater than standard glasswool insulation, so a wall with R2.5 batts will perform closer to R2.0.

Figure 3: Thermal conductivity of typical materials

By far the best thermal conductor in typical home construction is steel. We tried to minimise the use of steel, but where it absolutely was required (the spine of the butterfly roof and as supports for the chimneys) we used phenolic foam sheets to try and keep the steel on the cool (outer) side of the insulation.

Figure 4: Thermal bridging detailing

6.2 Thermal imaging

We found thermal imaging very useful to identify weaknesses in the insulation installation, particularly around corners where it was not evident to the naked eye the insulation was incomplete. Examples are shown below.

7 Airtightness

This is almost certainly the single hardest aspect of the passivhaus method to achieve. Extreme attention to detail is required to achieve the passivhaus levels of airtightness.

Airtightness is expressed as the air leakage volume per hour divided by the building volume at 50 Pa of pressure difference (usually called air changes per hour, or ACH50 or simply n50). The passivhaus standard uses the conditioned building volume; that is, the volume between internal linings and excluding internal walls. So, for example, a value of 1.0 means that at 50 Pa pressure difference the equivalent of one internal building volume of air would move in or out of the building over an hour. In practice 50 Pa is a very large pressure difference and would be expected to occur only on windy days, and even then only on buildings on exposed sites. In typical conditions a house of this level of airtightness may leak 5 to 10% of the building volume per house.

For new builds the passivhaus standard requires 0.6ACH50 and for retrofits 1.0ACH50. The typical new Australian home has around 15ACH50 and older homes, similar to our home before renovation, achieve levels up around 30ACH50. After renovation we achieved 1.2ACH50, or around 13 times better than the average new Australian home and just short of the EnerPHit target of 1.0ACH50.

Figure 5: Airtightness of typical homes

7.1 Airtightness design approach

Achieving passivhaus levels of airtightness is very challenging, particularly on a retrofit. The levels of airtightness required are far beyond what can be achieved using seals arounds doors and windows, and caulking around the skirting boards. Instead, a continuous membrane across the wall and ceilings was required. This membrane which was taped and glued to windows and at all joins. Moreover, there needed to be a minimum of penetrations of the membrane and where it was penetrated tapes and glues were used to ensure an airtight seal. In a building of this level of airtightness there are no bathroom extraction fans and the cooktop rangehood is internally vented (with a carbon filter). Moreover, any penetrations for plumbing and cabling must be minimised and sealed.

Figure 6: Airtightness detailing

7.2 Measuring airtightness

Airtightness is measured using a blower door test. In this method the building is fitted with a fan, usually in a door opening, and the building is both pressurised and depressurised. This is an incredibly insightful test; we very quickly found issues during our testing which would never otherwise become evident.

Figure 7: Blower door testing

The first test occurred after the membrane was installed but before the internal linings were added. The initial test result came in at around 7ACH50, which is barely better than a well-built conventional home. It was readily apparent there were several large areas (10 - 20 cm) where the membrane had not been taped. Once these were sealed the test result rapidly improved to around 3ACH50. Over the following week visual inspection and fixes produced a result of 2.2ACH50, which then improved over the course of a days’ work with the home pressurised and five people to 1.2ACH50. A third test several days later produced only an incremental improvement from 1.18 to 1.13ACH50. Finally, after internal linings were installed and the home was functionally complete 1.20ACH50 was achieved. The slight deterioration in performance is largely attributable to the reduction in internal volume that accompanied the installation of the internal linings. The key messages here are that:

  • even with due care and attention it is easy to overlook very substantial gaps,
  • blower door testing is essential to identify these gaps, and
  • it becomes very difficult at extreme levels of airtightness to identify small and diffuse gaps.

Figure 8: Improvement achieved on airtightness during construction

7.3 Mechanical ventilation

When airtightness drops below about 3.0ACH50 mechanical ventilation of the building is necessary to maintain a healthy air quality. The ventilation system performs two functions:

  • it provides constant, regulated and filtered fresh air, and
  • it recovers heat from exhaust air and uses this to heat incoming cool air (in winter).

The ventilation system runs constantly and uses about 50W, which is tiny in the context of the total home electrical load.

Figure 9: Mechanical ventilation and heat recovery unit

Figure 10: MVHR ductwork

Figure 11: Duct temperatures in MVHR across a typical day

It may seem incongruous that an energy efficient home would require the complexity, cost and additional electrical load associated with mechanical ventilation. However, in a conventional home we rely upon poor construction practices (i.e. leaks) to naturally ventilate our homes. We have no means to control this leakage, aside from opening windows, and it will vary with wind conditions in a way which is entirely beyond our control. Mechanical ventilation redresses all of these shortcomings:

  • it provides a definitive, controlled regulation of fresh air into the home, and
  • it captures the heat in the air that is being extracted and retains it within the home.

Moreover, there are secondary benefits - for example, the air coming into the building is filtered so there is reduced risk of allergies and less dust to move around the home.

8 Electricity

8.1 Solar power

We have 6.4 kW of LG NeON2 solar panels and SolarEdge DC optimisers with a Solar Edge 5k inverter. This system provides far more electricity than we require on a typical summers’ day, and around a third of what we require in winter.

8.1.1 Oversizing

The panels are oversized relative to the inverter (i.e. 6.4 kW of panels on a 5 kW inverter). This results in some clipping in the summer, but only reduces generation by around 1.3% per year. Most grid operators, including ours, limit exports to 5 kW. However, for the vast majority of the time the panels are not operating anywhere near their rated capacity. Moreover, the cost of panels is comparatively cheap so it makes sense to oversize in order to generate as much power across the year as possible; the only limitation being that the CEC restrict oversizing to 133% of the rated inverter capacity.

8.1.2 Panel orientation and shading

We suffer from some panel shading during the morning in particular, plus we had to orient the panels across two directions (north and west). A conventional string inverter would struggle to perform well under these conditions. Instead, we used SolarEdge DC optimisers on each panel. This setup allows each panel to operate to their optimum efficiency unaffected by shading on adjacent panels. Doing so is slightly more expensive than a string inverter, but in our case was essential. Another bonus is that this way we can monitor the performance of each individual panel, so it is easy to verify all panels are performing adequately.

8.2 Typical electricity consumption

Compared to many homes we have comparatively high electricity consumption (15 - 20 kWh/day in summer and closer to 30 kWh in winter). This is in part due to the sole use of electricity for providing hot water and heating and the electric car. The latter explains around a third of our winter consumption and half of our summer consumption. Moreover, car charging tends to happen in the evenings when the solar power is not generating (Figure 12: ).

Figure 12: Electricity profiles for typical summer and winter weekdays

The solar PV was only fully metered in late February 2017. Over March and April we tended to export around twice the electricity that we imported from the grid (Figure 13: ). This situation was reversed in the winter months such that so far we have imported 12% more electricity than we have exported. We hope over the full 12 month cycle we will be energy positive, or close to it.

Figure 13: Daily electricity import and export

Total import: 4,356 kWh
Total export: 3,905 kWh
Total generation: 5,219 kWh
Total consumption: 5,670 kWh
Solar self-consumption: 25%

9 Heating and cooling

9.1 No gas!

The home operates without gas; one of the first things we did was to disconnect the gas and ask the gas network operator (Multinet) to cap the gas line in the street. Instead, we have an electric oven and induction cooktop and use heat pumps for hot water and space heating and cooling.

9.2 Hot water and hydronic heating: air source heat pump

The hot water system consists of two Tivok air source heat pumps and a 420L storage tank provided by Enter-Shop and installed by Solar Hydronics. The heat pumps work in exactly the same manner as a conventional split-system air conditioner, but rather than heating or cooling air they heat water. The primary benefit of the heat pumps is that they operate at more than 100% efficiency. By putting in 1 kW of electricity the heat pump generates around 3.5 kW of heat energy by extracting heat from the air. The ratio of output to input energy is the coefficient of performance (COP). The Tivok has a COP of around 3 at typical winter temperatures. This is significantly lower than CO2 based systems such as the Sanden heat pump. While we’d have preferred to use the latter it is no longer warrantied for combined hot water and hydronic heating systems.

Figure 14: Coefficient of performance of typical air source heat pumps (note: test conditions vary, so comparisons are indicative only)

Figure 15: Air source heat pump and heating equipment

The effective heat output at winter temperatures from the heat pump is around 5.7 kW. While the supplier convinced us to buy two heat pumps we’ve only needed one in practice, even in winter when the heat pump must provide both hot water for showering and feed the hydronic system.

9.3 Heating

Heating is primarily supplied by a 2 kW hydronic floor coil in a 80 mm screed (Figure 16: ). The screed is insulated from the structural slab underneath by 80 mm of Kingspan Kooltherm phenolic foam (R4.0). The floor coil is complemented by two heated towel rails, one in each bathroom, each of 0.7 kW. This gives a total heat input into the building of around 3.4 kW. The combination of the high levels of insulation and the MVHR means there is minimal temperature variation between rooms.

Figure 16: Heating systems

In practice the heat input is a little undersized for cooler days, particularly when overcast. Our plan is to add a couple of radiators to bring the heat output closer to the single heat pump capacity of just under 6 kW.

One of the observations we have is that the absence of drafts, due both to the absence of air infiltration and from the high surface temperatures which preclude convective air movement, means that we feel comfortable at air temperatures lower than would otherwise be the case. For example, while our previous home thermostat was set to 21C in this home temperatures of around 19C feel similarly comfortable (or even moreso).

9.3.1 Insulation of hydronic piping

The hydronic PEX piping came fitted with 6 mm of foam insulation, giving a heat loss of around 15 W/m at 50C water temperature and 10C ambient air temperature. Without insulation the heat loss is around 430 W/m. After installation the last metre or so along the pipes was left uninsulated (Figure 17: ). By our calculations the total line losses in the circuit amounted to 1.2 kW or so, which is a significant proportion of the total heat load on the radiators (3.4 kW). By insulating these uninsulated sections, and also adding a second 13 mm foam insulation layer to the already-insulated sections we estimate a reduction in line losses to around 0.32 kW. Retrofitting insulation in an enclosed underfloor area is hard, clearly it would have been better to improve the pipe insulation prior to installation.

Figure 17: Hydronic heating pipes

9.4 Cooling

The combination of the insulation and airtightness helps to reduce heat gains in summer. Moreover, the glass has a low solar heat gain coefficient (SHGC=0.31) which reduces the solar heat gains in summer (this works against us in winter though). The eaves have been sized to prevent the midday summer sun from striking the windows directly, and there is an external blind on one window. We’re growing deciduous trees and climbers to protect other windows, mainly those facing west.

When required we’ve also got a Daikin Cora 5 kW air conditioner to provide additional cooling (this also provides additional heating in winter when required).

10 Rammed earth

Part of the walls of the home are stabilised rammed earth. This is a mix of gravel and sand with around 10% cement to act as a binder. The walls are 300 mm thick and consist of around 32 m3 of material, providing a high thermal mass. The rammed earth is structural; it supports the roof structure. However, rammed earth is a very poor insulator (\(\lambda=1.25 W/mK\)). To redress this we put an external stud wall (offset by 30 mm from the rammed earth) and used 80mm of Kingspan Kooltherm (a phenolic foam with a very high thermal resistance) to insulate the wall. In addition, the wall sits on 50 mm of high compression extruded polystyrene (XPS) to thermally decouple the wall from the footings and ground. In this way the intention is that the rammed earth can absorb and retain heat within the building thermal envelope.

Figure 18: Rammed earth

11 Contractors

Building designer EME Design
Builder Ridge Developments
Landscaping Haydn Barling Landscapes
Mechanical ventilation
- System supplier Fantech
- Installer Arvio
Rammed earth Earth Structures
Hot water heat pumps Enter-Shop
Blower door testing Efficiency Matrix
Solar PV Glen Clark & Co