Thursday, June 22, 2006

Comparison of Multifoil Insulation vs Celotex in a Model Experiment


I am in the process of a having a barn converted and have been intrigued and also highly frustrated by the debate on the pros and cons of multifoil insulation material. The builder who is leading the work on my barn conversion suggested we consider using multifoil insulation about a year ago - it was very easy to apply, more expensive but much more efficient. It sounded very promising and ideal for a wooden barn conversion where there was little space and we wanted to expose as much of the timber work as possible. I became aware of the differences of opinion about this new material after doing a quick search on the web, but I put it down primarily to competitor infighting, not uncommon with a new technology.

As a physical chemist I could see the theoretical benefits of using reflective materials to reduce heat loss and was very tempted to go ahead, despite there be no conclusive evidence of it’s effectiveness. The new insulation material clearly has a potential advantage for barn conversions and I was eager to get on with it - that was until I read the discussion on Multifoil insulation on the green building forum The posting by "Biff" on his kitchen experiment to measure the insulation properties of foil vs bubble wrap and a few other materials, left me with some doubts as to the insulation benefits of the multifoil. This together with some of the arguments by the different suppliers led me to carry out my own experiments which I hoped would give me the reassurance that the foil insulation was the way to go. I therefore set up an experimental model to simulate heat loss from a sealed container in which only the insulation material was varied.

The suppliers of the particular multifoil material claimed that it was equivalent to 100ml of celotex. Despite this claim, I chose to compare the foil with a layer of 50mm celotex, expecting to achieve a superior result.

Experimental Approach

Details of the various setups are shown in figure 1.

A 5 litre plastic container filled with hot water was centrally located inside a similar shaped 10 litre plastic container and insulated with either celotex or multifoil. The smaller container was isolated from the larger container by wooden spacers, creating an air gap of c. 20mm between the two containers. After the hot water was placed in the inner container, a 50mm thick tightly fitting lid was put in place and the temperature recorded with a digital thermometer with the sensing tip located c. 20mm below the surface of the water. The temperature of the hot water was monitored over a period of 5 or 6 hours and in 2 experiments overnight.

The lid and base support for the experiments were always the same and made from Celotex. This was purely for convenience since there was a risk that heat loss round the thermometer area could have been different for the different materials. Keeping the lid constant avoided this complication. The principle variable was the insulation material around the sides and bottom of the 10 litre plastic container.

After a number of feasibility experiments 4 controlled experiments were carried out.

1. No insulation material - The 10 litre container was simply encased in a cardboard box (the same box as used to support the multifoil in a later experiment). The top and base was Celotex as in all subsequent experiments, but there was no insulation material around the sides.

2. Multifoil insulation – A layer of a propriety multifoil insulation material (20mm thick, 14 layers of foil) was wrapped around the carton box as used above. The edges of the foil were overlapped by about 5cm. and sealed with "duck tape" (Figure 2)

3. Celotex insulation - A configuration in which the 10litre plastic container was encased by a Celotex box fabricated from 50mm Celotex board material. All joints were sealed with duck tape(Figure 3) In this configuration, no carton box was used.

4. Multifoil Insulation repeat - A repeat of exp. 2 with multifoil but to avoid possible effects of crimping of the foil by the tape, the insulation material was sealed without any compression by simply “butt jointing edges using using silicone sealant. As far as possible the insulation material retained a constant thickness of c. 20mm.

For each of the experiments water at approximately 80C was poured into the container, the lid and thermometer inserted and the temperature monitored over a period of 4 to 5 hours. In both experiments 3 and 4 the temperature was also measured the next day (about 18 hours after set up).

In addition to measuring the temperature of the water, the external temperature was also periodically checked. The ambient temperature varied from 11C to 18C depending on the time of day. By using a high initial water temperature, the effect of external temperature could be minimized. Furthermore In experiments 3 and 4 the ambient temperature variation was almost identical.


A summary or the experiment results is given in the graphs (figures 4 and 5). The starting temperatures varied slightly and therefore the data is adjusted along the time axis so that all graphs intersect at a common temperature for ease of visual comparison. Figure 4 showing the detailed temperature change in the first 4 to 5 hours and figure 5 showing that these differences are maintained over an 18 hour period for experiments 3 and 4.

All curves show a logarithmic decline in temperature as one would expect for a heat loss experiment. However, there is a slight deviation for logarithmic behavior in the first 15 to 20 minutes of setting up the experiment. This is attributed to the time it takes for the heat flux to equilibrate. The temperature of all materials needs to come to equilibrium and this leads to a slightly higher apparent heat loss in the first few minutes. After everything has settled down then the curve is logarithmic. For the technically minded this is demonstrated in figure 6, in which the logarithm of T-To ( ambient temp) vs time is clearly linear.

The data is quite conclusive - the multifoil insulation is not as good at retaining heat as the 50mm celotex although clearly it is better than no insulation at all. The repeat experiment for the multifoil insulation with a slightly different sealing approach gives almost exactly the same results. The foil behaves almost 50% worse than the celotex despite claims to the contrary. The time for the system to drop by a fixed temperature is nearly twice as long for celotex vs multifoil foil.


This set of experiments was designed to compare the two insulation materials at heat retention under as near identical conditions as possible. My objective was to provide data which I could use to make a simple decision on which material to use for insulation of my barn. Despite repeat experiments under near ideal conditions the Multifoil insulation was clearly inferior to the celotex. Although I was disappointed with the outcome, because of the practical consequences, the decision on which sort of insulation material to use was obvious.

In hindsight I am not surprised by the result, since if the foils claimed benefit is due to it’s reflective properties, then logically only the first layer will have any real effect. Any subsequent layer reflects heat back into the material itself – this heat is trapped and the foil heats up. Heat loss would then be by conduction along the various sheets of aluminium foil.

The controversy of multifoil as an insulation material will no doubt continue in the building trade. However, as a simple consumer and end user of the building it has to protect, I am sorry to say that multifoil insulation would not be my first choice. Only if space was limited and the structural configuration complex, would I consider multifoil as an alternative, but then as a last resort. It is just not cost effective compared to other materials.

Fig 1a - Experimental Set up.

Fig 1b Plastic Containers used to provide the heat source and the air gap

Fig 1c Plastic Containers Assembled with lid and Digital thermometer

Fig 2 - Multifoil Insulation Set Up

Fig 3 Celotex Insulation Set Up

Fig 4 - Graph of Temperature vs Time 6 hours duration

Fig 5 - As for fig. 4 including 18hour data for 2 experiments

Fig. 6 Logarithmic graph for T-To vs Time