Straw bale building is constantly developing in Europe and elsewhere. More than a thousand buildings, from dwellings to public buildings, were enumerated, as discussed, in the recent European Straw Bale Gathering [
1]. The straw bale community is well-organised and straw bale building techniques are still evolving. Among new developments in prefabrication, building certification, fire resistance, and moisture transfer validations, the thermal behaviour of straw bale walls is one of the key topics in market development.
Despite the straw variability inherent to vegetal materials, the heterogeneous nature of this efficient insulation material is not seen as an obstacle to its development. A rational analysis of the parameters that influence its thermal conductivity can support a rigorous assessment of the energetic and environmental benefits of choosing straw bale building techniques. Straw bales have a larger thickness than most of the insulating materials that can be found on the building market. Measurement apparatus for thermal conductivity is usually not designed for such thickness, and most of the thermal conductivity values that can be found in the literature are defined based on samples in which straw bales are resized according to standard EN ISO 10456 [
2] . During this operation, the orientation of the fibres and the density may not be preserved. A specific Guarded Hot Plate was designed in Belgium [
3] to measure samples up to 50 cm thick, according to the reference standard ISO 8302 [
4].
This paper starts with a literature review of straw bale thermal conductivity measurements and presents a measuring campaign performed with this apparatus. The influence of each analysed parameter is discussed thoroughly. Representative values are proposed for a large range of straw bales to support straw bale development in the building industry.
1.1. Overview of Thermal Conductivity Measurements on Straw Bales
In most of the available literature, two distinct values for the thermal conductivity of straw bales can be found. The first one gives the thermal conductivity when the straw fibres are perpendicular to the heat flow. The second one gives the measured value when the fibres are parallel to the heat flow.
The next paragraph focuses on this specific aspect. It must be noted that this distinction was set up by McCabe [
5] in one of the first well-referenced works relating to straw bale thermal conductivity. In this work, McCabe showed, based on measurements made with a device akin to a hot plate, that the thermal conductivity of the straw bales is 0.048 W/mK when the heat flow is perpendicular to the fibers and 0.061 W/mK when parallel.
After this work was published, many similar experiments were conducted, such as that performed by as Andersen [
6] and Shea [
7]. Andersen studied the influence of the density on the thermal conductivity. They studied two sets of samples, one with a density of 75 kg/m
3 and the other with a density of 90 kg/m
3. The thermal conductivity of the first set of samples was 0.052 W/mK when measured perpendicular to the fibres and 0.056 W/mK when parallel to the fibers. The thermal conductivity obtained for the second set of samples was slightly higher; 0.056 W/mK when perpendicular and 0.06 W/mK when parallel. These values point out the first result regarding the evolution of the thermal conductivity with the studied parameters. When the density is 75 kg/m
3 and the investigated direction evolves from perpendicular to parallel, the measured thermal conductivity increases by 0.004 W/mK; for a 90 kg/m
3 density and measurement directions evolving from perpendicular to parallel, the same variation of magnitude is observed for the thermal conductivity. Consequently, the thermal flux measured by these authors can be modeled considering the density and the flux direction independently.
Shea studied the thermal conductivity with a flow meter on various samples of straw with a thickness of 25 cm, compressed to reach a large range of densities. In this particular experiment, the straw fibers in the samples did not have a specific orientation. The measured thermal conductivities ranged from 0.059 W/mK for a density of 63 kg/m
3 to 0.064 W/mK for a density of 123 kg/m
3. Shea proposed a reference value of 0.065 W/mK. FASBA, the German association for straw bale buildings, led numerous researches on straw bales and obtained in 2010 very good thermal conductivity, around 0.045 W/mK, when the heat flow was perpendicular to the fibers [
8].
Many other data on thermal conductivities can be found in the literature. Among them, the value validated by the German Centre of Competence for Construction (DIB) [
9] is often used as a reference in many countries. The thermal conductivity is there considered to be 0.052 W/mK, when the heat flow is perpendicular to the straw fibers, and 0.080 W/mK when it is parallel. Although reliable measurements are usually obtained using steady state methods such as the ‘guarded hot plate technique’, Dubois [
3] underlined that most of the data derives either from transient methods or from steady methods with low thickness rebuilt bales. Douzane [
10] presents the results of thermal conductivity measured on straw bales, obtained through a steady state method. A guarded hot plate apparatus was used to evaluate the thermal conductivity of the straw bales. As the apparatus was a commercial device, the samples were prepared by cutting off straw bales to low thicknesses (10 cm). Two kinds of samples were investigated in relation to the orientation of fibers. The average values of the thermal conductivity at 10 °C were, respectively, 0.072 W/mK and 0.051 W/mK in parallel and perpendicular orientations. Conti [
11] developed a thermal conductivity measurement system for straw bales that was also based on a steady-state method. The designed hot-box is mainly composed of a metering chamber and a heater inside a climate chamber. Authors found a thermal conductivity around 0.066 W/mK in the case in which the thermal flux was considered parallel to the fibers.
The effect of relative humidity on straw conductivity is found to be significant by several authors; according to Wei [
12], conductivity increases with moisture content due to the porous structure of the fibrous insulation materials. Wei measured an increase of the thermal conductivity of rice straw when the moisture content was varied; 0.0514 W/mK for 10% of moisture content and 0.0519 W/mK for 18% of moisture content. The same results were obtained by Grelat [
13] for straw bales. Measurements with 0%, 50%, and 90% relative humidity (0%, 15%, and 22% water content) showed a significant increase of the conductivity; from 0.064 W/mK for 0% RH to 0.069 W/mK for 22% RH. Palumbo [
14] investigated the thermal conductivity of a board composed of barley straw (81%) and corn starch (19%). Observations also lead to a significant linear increase of the thermal conductivity when the relative humidity varied from 10% to 90%.
Table 1 gives a synthetic view of these results. The
Figure 1 and
Figure 2 display the distribution of the thermal conductivity values found in the literature, respectively along with the density of the bales and the direction of the thermal flow.
Some immediate conclusions can be brought out; considering a given thermal flow direction, these observations show a rise in thermal conductivity when the density is increasing. When the direction of the heat flow is considered, the measured conductivity values show a decrease when the direction is varying from parallel to perpendicular. The case observed by Shea (no specific direction—represented by a blue circle in
Figure 1) can be considered as a mix of parallel and perpendicular fiber directions. Consequently, it was chosen to display the values along the x-axis in
Figure 2, between the parallel and perpendicular directions.
If many values are given for perpendicular or parallel heat flow directions, not a lot of information can be found on how the fibers are truly oriented in the straw bales.
Figure 3 shows the average results obtained when analyzing five slices of a single straw bale. This visual assessment accounted for each slice; 50 fibers of a minimum 8 cm long.
As presented in Shea’s paper [
7], it can be observed that, in the studied straw bales, the straw fibers appear to be randomly oriented. Following the directions defined in
Figure 3 and detailed in
Section 1.2, the results presented in
Figure 4 show that a slight majority of fibers oriented toward direction 2 can be observed.
This observation goes against the belief that the thermal conductivity of straw bales is lower when the bales are laid on edge.
In addition, due to the way the straw is processed in the balers, one can easily observe that, if the heat flow follows direction 3, the fibers can globally be assumed to be perpendicular to it.
1.2. Straw-Bale Orientation in Vertical Walls
To clarify what is meant by referring to straw bales ‘laid on edge’ or ‘laid flat’, it is proposed to name the surface of the straw bale that can be seen when the bale is installed in the wall, based on the heat flow direction.
Figure 4 illustrates the three directions of heat flow. If, in a vertical wall, the bale is laid ‘flat’, the heat flow follows direction 1. If the bale is laid ‘on-edge’, the heat flow follows direction 2. Direction 3 is almost never encountered in building assemblies.
‘Surface 1’ (36 cm × 80 cm) is thus defined to be the surface of the bale seen when the bale is laid ‘flat’ and where the heat flow will follow direction 1. ‘Surface 2’ (46 cm × 80 cm) is the surface that can be seen when the bale is laid ‘on-edge’ and where the heat flow will follow direction 2. ‘Surface 3’ (36 cm × 46 cm) is the ‘header’ of the straw bale, which can be seen if the bale is laid with its long side perpendicular to the wall surface, and where the heat flow follows direction 3.
Figure 5 illustrates these definitions.
In many cases, e.g. load bearing and GREB techniques, the straw bales are laid flat like most classic brick or stone walls, as illustrated in
Figure 6. The bales are then lying on surface 2, which is horizontal in this case. Surface 1 can be seen when the wall is completed. The insulation layer is 46 cm in this case, and, due to the way the bales are processed in the balers, surface 1 is a good substrate for rendering and plastering (with good mechanical cling). The heat flow follows direction 1. If there is a load bearing structure, the minimum spacing between the vertical posts is 80 cm.
Figure 7 shows a variant where the straw bales are laid on surface 3. Again, the insulation layer is 46 cm thick and the heat flow follows direction 1. The surface of the wall (surface 1) is a good substrate for rendering and plastering. The minimum spacing between the vertical posts of a load bearing structure is 36 cm.
The bale can also be laid ‘on-edge’ (i.e., on surface 1). The insulation layer is then 36 cm thick and the heat flow follows direction 2. This position is sometime used in the GREB technique. It is usually chosen to avoid walls that are considered to be too thick. The minimum spacing between the vertical posts of a load bearing structure is here 80 cm. The variant presented in
Figure 8 illustrates another wall with a 36 cm thickness, in which the straw bales are laid on surface 3 and where the heat flow follows direction 2. The minimum spacing between the vertical posts of a load bearing structure is thus here 46 cm.
Figure 9a,b present a variant in which the straw bales are laid on surface 2. This orientation is not common on building sites. This theoretical orientation of straw bales in the wall is presented because it may have an interesting thermal performance, as the results on thermal conductivity measurements will show. In this case, the straw bales are laid on surface 2 and the heat flow follows direction 3. The
Figure 9a shows that, without reducing the length of the straw bales, the wall would be 80 cm thick. The
Figure 9b shows a more realistic proposal where the straw bales are produced (or adapted) to obtain a wall about 36 cm thick. The minimum spacing between the vertical posts of a load bearing structure is then 36 cm. Specific attention must be paid in these two cases if a rendering or plastering is to be applied on the wall because surface 3 may not be a good substrate (bad mechanical cling).
These few examples show the diversity of building techniques that can be designed when using strawbale walls. It also shows that the position and orientation of straw bales in the wall will influence the direction of the heat flow within the bales.