I guess that the above title was never used because convection (forced and natural) in mineral wool is not clearly communicated. For example, on the Rockwool website, we read
ROCKWOOL stone wool achieves its insulating properties by ‘capturing’ the air between the fibers, so that virtually no convection takes place. Because ROCKWOOL insulation only contains natural air and no other gases such as blowing agents, the thermal performance does not change due to gases diffusing from the products – not even if longer periods like the total lifetime of a building are considered.
Virtually no convection is rather vague while we read on the Paroc website :
Heated air becomes less dense and rises and cooler air is drawn in to fill the space left by the displaced heated air. Natural convection might occur, for example, in a very low-density mineral wool insulation layer during extremely cold winter days.
Paroc writes clearly something different than Rockwool, while they are producing the same product. “Virtually not” and “during extremely cold winter days” are for me two different things.
Like already discussed, Vacuum Insulation Panels (VIP) can be produced from low density mineral wool and it is observed that the thermal conductivity decreases much more due to the vacuum than expected if we assume that the air is still (no natural convection).
The circumstances, where natural convection will be present and may even double the heat transfer as expected from the “laboratory” thermal conductivity, are more or less defined by the modified Rayleigh number. Modified means that the permeability of the insulation for air is included. This permeability is a very important parameter in this case and it is rather well known for Rockwool thermal insulation. Like can be observed on the graph, it is not linearly related to the density. We developed a spreadsheet convection where the modified Rayleigh number is calculated and where the reader can calculate how much the heat transfer is increased compared to the assumed “virtually no convection” case. The spreadsheet is based on a paper of Paula Wahlgren of Chalmers University in Sweden. We use a measured permeability at 30 kg/m³ density and extrapolate with an air Flow Resistance graph. We also included the U-value of the thermal insulation. The calculation shows that:
- A passive roof in Denmark (U=0.06) with 30 kg/m³ mineral wool has the onset of natural convection at a temperature gradient of 40°C (outside = -20°C). Increasing the density (and so the cost) is the only way to obtain the U=0.06 with mineral wool.
- Industrial insulation of 80 kg/m³ is typcially used with a temperature gradient of 600°C and 35cm thickness in the glass industry in the assumption “virtually no convection). In reality, the heat transfer due to convection is an extra 30%. To reach the safe side, we need to increase the density to 140 kg/m³, which is almost a doubling of the cost.
In both cases, it is clear that we have to increase the density of the mineral wool up to the ones used in the cellular glass world.
Another paper, written by M. Serkitjis, Chalmers University in Sweden, demonstrates the effect of natural AND forced convection on low density mineral wool. It is shown that a draft above mineral wool lowers the thermal resistance of low density mineral wool. “If wind velocity above the insulation is as much as 1.5 m/s, which is reasonable in an attic space, heat losses can increase by 10 – 30% depending on the permeability of the material.”
Low density mineral wool may be a cheap non-combustible thermal insulation but can only be used in moderate circumstances. For passive housing with cold winters or under windy circumstances, an important extra thermal loss may show up. Higher densities are the only solution but in that case, cellular glass is a very interesting alternative. Cellular glass is an non-combustible convection free thermal insulation. GLAPOR cellular glass has comparable prices with mineral wool at equivalent prices. This is quite logic while GLAPOR cellular glass is foamed at 800°C and mineral wool involves a melting up to 1600°C.
Pingback: Real life investigation of foamed glass gravel | BELGLAS BVBA