“INSULATION PRODUCTS – A COMPARATIVE SURVEY”

a presentation by

Mr Roeland Ansems
Ansems Engineering and Management

A COMPARISON OF TYPES OF THERMAL INSULATION FOR BUILDINGS

The literature and marketplace is best with contradictory claims concerning relative strengths and weaknesses of various types of thermal insulation.

The need for valid comparison is particularly pressing in the light of impending legislation in several Australian states following the introduction of Mandatory thermal insulation standards for dwellings in Victoria in 1991.

To make a meaningful comparison of insulants we must consider following key criteria:

    1) Thermal Performance
    2) Cost
    3) Safety: Fire Performance
    4) Safety: Smoke Toxicity
    5) Overall Performance Comparison

1)    Thermal performance is the principal function of any thermal insulation.
Durability and quality assurance are also important aspects of thermal performance but they are beyond the scope of the present paper.

2)    Cost is a local commercial issue and one which is difficult to address in a single
sweep of the marketplace.
Nevertheless, we shall look at some of the key factors which should be taken into
account when assessing the cost of thermal insulation.

3,4) The question of Safety is a very broad issue and under this heading we propose to examine only Fire Performance and the associated issue of Toxicity of Smoke evolved in event of fire. These safety factors are key objectives of the Building Code of Australia and therefore must be strictly controlled.

1. THERMAL PERFORMANCE

This property is specified nowadays in Australia under two possible headings:

    i)  Thermal Conductivity ("k", w/mºK)
    ii) Thermal Resistance ("R", m2 ºK/w)

Thermal conductivity (k) value is relevant if building space or cavity thickness is limited.

Thermal Resistance ( R ) value is relevant in those situations where a degree of freedom is available regarding thickness (e.g. in attic/roof spaces of most domestic dwellings).

The architect or specifier must be aware of the relationship between k and R for homogeneous solids.

    R = l/k

Where 1 is the actual installed thickness of the insulant.

Note that data sheets for some compressible insulants will show R value at some theoretical "Test Thickness" which may bear no relation to the thickness of the insulant in situ.

This factor is of particular concern for those products which sag, settle or otherwise deteriorate over a long period of time, since this will cause loss of thermal performance.

K, thermal conductivity, depends on Temperature.

It is important to ensure that the conductivities (or R values) of various insulants to be compared are measured at the same mean temperature (e.g. 25ºC in Australian buildings).

There is an increasing number of new insulation candidates which have no established Australian Standards and then it is a case of "Caveat Emptor" - Let the Buyer Beware!

Thermal performance of a range of products is categorised in the following table.

Note: Data for "foil batts" are included for the winter case only (heat flow upwards in roof space).

Further research (perhaps involving a full in-situ thermal performance study of the reflective batt system) is advocated in order to verify thermal performance claims of foil based systems. Some foil manufacturers' claims appear unrealistic and do not appear to take into account dust, conductive thermal bridge effects of the foil gussets in the batts, nor do they seem to adequately address the question of air circulation/convective heat gains or losses in the spaces between joists and batts if they are poorly fitted or damage in any way.



2. COST

Economic considerations of various insulants should include:

1)    Transport cost: the compressible types of insulant generally achieve significant savings over the rigid forms (except foils).

2)    Installation cost: the so-called "blow in" insulants (loose fill) are cheap and easy to install. However, they may suffer from lack of long term durability in terms of settling, wind, moisture, vermin resistance and leaching of fire retardants.

3)    Environmental considerations.

Some foam insulants require CFC gas to produce their cellular structure. The long term cost to the environment cannot be quantified, but nevertheless the negative impact on the ozone layer should be taken into account when considering use of such products.

Furthermore, slow diffusion (leakage) of the gas from the insulant's "closed" cells will adversely affect thermal performance over time, as air replaces the gas.

4)    An objective comparison of cost should be done on the basis of total installed cost per unit of Area, per unit of R value. ($/m2/{m2ºK/W} ).
In the attached table all of the insulants are compared on the basis of the same R value (2.5 m2 ºK/w) unless for some reason they cannot achieve this in the typical ceiling situation (e.g. foil batts in ceilings in winter, cellulose if screeded off to 100 mm joists, with 15-20% settling over time).

The relative cost comparison on this basis is listed as "Thermal Performance Ranking" in the attached table.

A column headed "Mass" compares the weight of each insulant required to do the same job (R 2.5 for 100 M2 of ceiling area) Note the major variations in mass required: mechanical design constraints clearly rule out some materials on this criterion alone and many of the heavier materials are used only for other specialised purposes, such as fire retardancy (e.g. perlite).



3. FIRE PERFORMANCE

Fire Test AS1530.3 is currently in widespread use in Australia to assess early fire hazard of materials used under the Australian Building Code in domestic and commercial buildings.

The test uses a vertically mounted radiant panel facing a vertical sample to try to simulate the early development of a fire in a building, and it was specifically designed for furnishings & internal wall lining materials. The effect of direct flame impingement is not assessed in the test, nor does it accommodate materials of varying ignitability, since radiant flux ceases to increase in the test at the point when the material ignites. This can and does lead to anomalous results.

Four indices are generated by the test:

Ignitability, Spread of Flame, Heat Evolved and Smoke evolved.
The Australian Building Code uses only two of the AS1530.3 indices, the

Spread of Flame Index and Smoke Developed Index.

The relevant sections of the ABC apply to public areas, theatres, halls, transport terminals, corridors, hallways where materials must have spread of flame index of 0 and smoke developed index of not more than 5.
Note that there may be particular statutory requirements in various States which override the Australian Building Code. Thus, in NSW Ordinance 70, Para 16.19 (12) states that for a building more than 25 m in height, "any external insulating material used on ductwork, chambers or pipework shall have a spread of flame index not greater than 0 and a smoke developed index of not greater than 3".

The application of AS153.3 fire test to thermoplastics and foils is clearly inappropriate. In fact, the original edition, AS A.30-1958, p10, section 3.1 specifically excluded such materials: "..Since the source of heat in the test is a radiator, the test is not suitable for the accurate classification of materials that have highly reflective surfaces or that melt and flow prior to ignition in the test""

The current method (AS1530.3 1989 version) does however require that reflective surfaces be blackened and scored with a knife prior to subjecting faced samples to radiant flux in the test.
There is increasing concern over the widespread misinterpretation of results from AS1530.3 which is currently being applied to almost every type of insulation considered here, including thermoplastics like polystyrene foam, polyethylene foam, dacron (polyester fibre), modacrylic fibre and even foil faced thermoplastic insulants and foil batts.
AS1530.3 is used even on complex composite systems such as compressible airconditioning duct under the requirements of AS1668 Mechanical Ventilation & Airconditioning Code.
AS1530.3 is currently under revision by SAA BD/18 and AUBRCC committee AP73.
Use of AS1530.3 (and part 2-Flammability) in the building code will probably be supplemented by other Fire test methods, such as ASTM E970 Attic Radiant Panel test for ceiling insulation, BS 476.7 Radiant Panel test, ASTM E84 Steiner 3.

Tunnel Test or the Canadian B54.9 Tunnel test for wall/ceiling panels and UBC 17-5 Corner Wall test for Polymers and foams.
For Cellulose Fibre, largely as a result of NSW Dept. of Consumer Affairs Product Safety Inquiry (1992) the ASTM C739 and BS5803.4 smouldering/inclined radiant panel fire tests are mooted to replace AS1530.3 in the current revision of Cellulosic Fibre Loose Fill Insulation Standard AS2462,1981, (amended 1985).

And for flexible ducting, the specially developed UL181 series of American Underwriters Laboratories standards and tests are likely to replace AS1530.3 in the commercial builing ventilation and airconditioning code (AS1668.1) and may also be adopted by SAA committees on airconditioning ductwork (ME/62.3).
It is clear that the AS1530.3 early fire hazard test was not intended to be used for foils or highly reflective surfaces; AS1530.2 is intended for these products.

Thermoplastic materials which melt and flow away from the radiant panel were excluded from the (original) scope of AS1530.03, and are better chracterised by, say, ASTM E970 Radiant Panel test or the American Corner wall test, UBC 17-5.
Nevertheless, it is well known that the judicious application of a thin layer of foil to one side of an otherwise highly flammable polymeric foam (e.g. polyethylene foam) or organic polymer fibre insulation product (e.g. Sheep's wool) may pass the AS1530.3 test with close to a "4 zero" rating, and thereby may obtain approval in commercial buildings.
Without the foil, sheeps wool typically gives 17,4,4,4 result in the AS1530.3 test.

Similarly, it is well known that the fire performance of any carpet or other textile product under Radiant Panel testing is highly dependent on the density, orientation and moisture content of the sample.
Ref. Handbook of Fiber Science & Technology, Vol II, Part B "Functional Finishes" Shaw & White, Ch. 5 "Wool Finishing".
Thus, wool carpets cannot pass ASTM E162-67 Radiant Panel Test, or BS476.1 (Dutch, Belgian, British high rise building regulations, at least class 2 required) unless Fire retardant is used on the wool and density of the wool exceeds 150 KG/m3.
In Australia, non-fire-retarded sheeps wool batts (density <10 kg/m3 and loose fill sheeps wool (density 32-40 kg/m3) are tested by AS1530.3 "held by a steel mesh screen of 12 mm pitch and 0.8 mm wire thickness, securely clamped in 4 places".

The wool sample density is increased substantially by this artificial procedure, and the surface fibres may be significantly flattened by the wire mesh.

Compression of any combustible sample excludes oxygen and may thus reduce its propensity to burn in terms of the test. In this way it is apparently possible to improve on the 17,4,4,4 result for sheeps wool sufficiently to pass the Australian Building Regulations (0.5 allowed for spread of flame and smoke developed indices).
A draft standard has been introduced for (loose fill) wool insulation and these issues will be dealt with during development of that standard.
It is quite clear that "compliance" with AS1530.3 will no longer be sufficient for many aspiring new building materials which may well present entirely unacceptable fire hazards in real fire situations.
A relative ranking in terms of increasing fire hazard of insulants is included in the attached table, based on a qualitative assessment of all the available test data and literature and not restricted only to AS1530.3 performance.

4. TOXICITY OF SMOKE EMISSIONS

AS 1530.3 attempts to evaluate only the visibility and flammability of gases evolved from a test sample.

It makes no attempt to assess the toxicity of the smoke/gases evolved in the fire situation.

Obscuration of vision is certainly a hazard, especially for the fire fighters who arrive after a fire is well alight.
However, toxicity of smoke evolved may not be indicated simply by smoke density but may have fatal consequences for the occupants of a burning building, particularly single-occupancy domestic dwellings in the absence of effective smoke detection devices.
Products of pyrolysis of some organic insulants may include CO2, H2O, NH3, CH4, H2S, CO, HCN etc. (Carbonyls)
The last three gases, Hydrogen sulphide, carbon monoxide and hydrogen cyanide, are colourless but highly poisonous and are of great concern due to their relatively low LC50 (dangerous to life in 30 minutes exposure) concentrations.
Assessment of relative toxicity risk of alternative insulation materials should determine how much insulation is present and how much toxic fume is emitted and at what rate burning occurs.

A suitable toxicity model might include:

1) k value or R value per unit thickness
2) Density in situ
3) Temperature of decomposition
4) Relative toxicity of combined offgases.

Toxicity data, such as LC50 or LT50 or LTC50
Numbers (empirically determined on mice in an exposure chamber) should ideally be determined in both flaming and smouldering combustion conditions, since some fire retarded insulations have a great propensity to smoulder for long periods of time.
Acute Lethal Hazard (ALH) may be defined as

    ALH = k x D / T / LC50

    Where     k = thermal conductivity,
            D = Density
            T is decomposition Temperature
            LC50 is the sample weight lethal to 50% of test animals.

   
Ref 2.
Alarie & Anderson, "Toxicologic Classification of Thermal Decomposition
Products of Synthetic and Natural Polymers", Toxicology & Applied Pharmacology, 57, 181-188 (1981)

Ref 3.
Alarie & Anderson, "Toxicologic & Acute Lethal Hazard Evaluation of Thermal
Decomposition Products of Synthetic and Natural Polymers".

Ref 4.
Hilado & Huttlinger, "Toxicity of Off Gases From Thermal Insulation
Materials: A Review" J.Thermal: Insulation. Vol. 5 (Oct 1981) p.73-77)

The attached table includes a listing of ALH values.
These are listed in increasing order of toxicity hazard.

5.    OVERALL INSULATION PERFORMANCE COMPARISON

An "overall performance ranking" may be derived as follows:
Po = {Pj + 2 (1+[ Pf + Pt ] ) } /3
Where

Po = Overall performance ranking (1-21)

Pj = thermal performance (price) ranking (1-21)

Pf = Fire performance ranking (0-10)

Pt = Toxicity (ALH) ranking (0-10)

The attached table lists values for this relative ranking index for all of the insulants studied.

6. CONCLUSION

For a series of commercially available thermal insulations we have considered a "standard" loading of insulation having the same Thermal Resistance for 100 m2 insulated area and R = 2.5 m2 ºK/W.

We have considered cost, relative fire performance, and fire toxicity, and have provided a simple points-based ranking scheme to give equal value to these factors.

In the case of fire performance, due to disparate test data, the ranking methodology is necessarily qualitative in nature.
But when combined with the thermal performance and toxicity data the overall performance index provides an incisive first-pass comparison of thermal insulation products for similar applications.
This ranking methodology may be a useful method of facilitating choices between the growing list of alternatives which are presented to the building designer.

Other properties might also be ranked for more extensive comparisons. Thus we might consider acoustics (NRC and STC data), mechanical properties, thermal mass (heat capacity), thermal expansion coefficients, permeability, environmental hazards, health and safety etc.
These are beyond the scope of the present discussion, which concentrates only on economics and those factors which are required in the Building Code of Australia. These are:

- Thermal Performance

- Fire Resistance

- Toxicity of Smoke evolved

* Note: Settling allowance for Cellulose 15-20% of installed thickness
* Note: R Value for Foil Batt for single cell winter hear flow upwards, "aged" dusty upper surface only no leakage losses or bridging allowed for.
# Note: Fire performance comparison based on best availabe data and is relative only since not all products tested by same test methods