Vacuum insulation panels for building applications: A review and ...

Review

Vacuum insulation panels for building applications: A review and beyond

Ruben Baetens a,b,c, Bjørn Petter Jelle a,b,*, Jan Vincent Thue b, Martin J. Tenpierik d, Steinar Grynning a,Sivert Uvsløkk a, Arild Gustavsen e

a Department of Building Materials and Structures, SINTEF Building and Infrastructure, NO-7465 Trondheim, Norwayb Department of Civil and Transport Engineering, Norwegian University of Science and Technology (NTNU), NO-7491 Trondheim, Norwayc Laboratory of Building Physics, Department of Civil Engineering, Catholic University of Leuven (KUL), BE-3001 Heverlee, Belgiumd Faculty of Architecture, Urbanism and Building Sciences, Delft University of Technology, Julianalaan 134, 2628 BL Delft, The Netherlandse Department of Architectural Design, History and Technology, Norwegian University of Science and Technology (NTNU), NO-7491 Trondheim, Norway

Energy and Buildings 42 (2010) 147–172

A R T I C L E I N F O

Article history:

Received 22 December 2008

Received in revised form 3 September 2009

Accepted 14 September 2009

Keywords:

Vacuum insulation panel (VIP)

Building insulation

Thermal bridge

Service life

Gas-filled panel (GFP)

Aerogel

Vacuum insulation material (VIM)

Nano insulation material (NIM)

A B S T R A C T

Vacuum insulation panels (VIPs) are regarded as one of the most promising high performance thermal

insulation solutions on the market today. Thermal performances three to six times better than still-air

are achieved by applying a vacuum to an encapsulated micro-porous material, resulting in a great

potential for combining the reduction of energy consumption in buildings with slim constructions.

However, thermal bridging due to the panel envelope and degradation of thermal performance through

time occurs with current technology. Furthermore, VIPs cannot be cut on site and the panels are fragile

towards damaging. These effects have to be taken into account for building applications as they may

diminish the overall usability and thermal performance.

This paper is as far as the authors know the first comprehensive review on VIPs. Properties,

requirements and possibilities of foil encapsulated VIPs for building applications are studied based on

available literature, emphasizing thermal bridging and degradation through time. An extension is made

towards gas-filled panels and aerogels, showing that other high performance thermal insulation

solutions do exist. Combining the technology of these solutions and others may lead to a new leap

forward. Feasible paths beyond VIPs are investigated and possibilities such as vacuum insulation

materials (VIMs) and nano insulation materials (NIMs) are proposed.

� 2009 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Energy and Buildings

journa l homepage: www.e lsev ier .com/ locate /enbui ld

Contents

1. Introduction to vacuum insulation panels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148

1.1. The vacuum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

1.2. The core. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

1.2.1. Physical properties of fumed silica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

1.2.2. Water vapour adsorption of fumed silica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

1.2.3. Thermal conductivity of fumed silica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

1.2.4. Fire behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

1.3. The envelope. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

1.4. Getters, desiccants and opacifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

2. Thermal bridges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

2.1. Thermal bridge effect on the scale of VIPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

2.1.1. Thermal bridging due to VIP envelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

2.1.2. Thermal bridging due to air gaps between two adjacent envelopes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

2.2. Thermal bridge effect on the scale of a building component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

2.2.1. Analytical model of Tenpierik et al. [30] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

2.2.2. Numerical models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

2.3. Thermal bridge effect on the scale of the building . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

* Corresponding author at: Department of Building Materials and Structures, SINTEF Building and Infrastructure, NO-7465 Trondheim, Norway. Tel.: +47 73 593377;

fax: +47 73 593380.

E-mail address: [email protected] (B.P. Jelle).

0378-7788/$ – see front matter � 2009 Elsevier B.V. All rights reserved.

doi:10.1016/j.enbuild.2009.09.005

mailto:[email protected]

http://www.sciencedirect.com/science/journal/03787788

http://dx.doi.org/10.1016/j.enbuild.2009.09.005

R. Baetens et al. / Energy and Buildings 42 (2010) 147–172148

3. Service life prediction for VIPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

3.1. Service life definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

3.2. Thermal conductivity of the core as function of moisture content and air pressure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

3.3. Pressure and moisture content increase as function of envelope material properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

3.3.1. Dependency of WVTR and GTR on panel size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

3.3.2. Dependency of WVTR on relative humidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

3.3.3. Dependency of WVTR and GTR on temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

3.3.4. Measuring the internal pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

3.4. Service life prediction for VIPs with a fumed silica core and foil envelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

3.5. Other ageing factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

4. Acoustical properties of applied VIPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

4.1. Acoustical properties of a single VIP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

4.2. Acoustical properties of VIP insulated structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

4.2.1. Vacuum insulated sandwiches VISs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

4.2.2. VIP insulated cavity wall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

5. Building applications of VIPs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

6. Other possible high performance thermal insulating materials and solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

6.1. Gas-filled panels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

6.2. Aerogels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166

7. Beyond vacuum insulation panels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166

7.1. Possible improvements for current VIPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166

7.2. New high performance thermal insulation materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

7.3. Vacuum insulation materials (VIMs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

7.4. Nano insulation materials (NIMs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168

7.5. Gas insulation materials (GIMs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

7.6. Dynamic insulation materials (DIMs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

7.6.1. Dynamic vacuum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

7.6.2. Dynamic emissivity of inner pore surfaces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

7.6.3. Dynamic solid core thermal conductivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

7.7. New thoughts and ideas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

8. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

1. Introduction to vacuum insulation panels

Current vacuum-based insulating solutions can be subdividedinto three categories: Vacuum insulation panels (VIPs), vacuuminsulating sandwiches (VISs) or sheet-encapsulated vacuum insula-tion panels and vacuum insulating glazing (VIG) (see Fig. 1). Thisreview deals about VIPs as a high performance thermal insulatingsolution for building envelopes [1]. Vacuum insulation panels can dedefined as ‘‘an evacuated foil-encapsulated open porous material as a

high performance thermal insulating material’’.The physics of the total heat transfer through an insulation

material are well known. The total density of heat flow rate in suchporous materials can be divided in four different heat transferprocesses: Heat transfer qr [W/m2] via radiation, heat transfer qcd

via conduction of the solid skeleton of the core and heat transfer viathe gas inside the material. This last mentioned transfer via theinternal gasses can be divided in heat transfer qg due to gas

Fig. 1. [Left] Vacuum technology as building insulation: VIPs, VIG and VISs [105] and [rig

with the same overall thermal performance [67].

conduction and heat transfer qcv due to gas convection. Thetotal density of heat flow rate qtot can than be approximated by thesum of the densities of these different heat transfer mechanisms[2]:

qtot ¼ qr þ qcd þ qg þ qcvðþqcouplingÞ (1)

The term qcoupling in Eq. (1) has to be added for powder and fibrematerials, for which the total heat transfer will be larger than thesum of the separated heat transfer mechanisms due to interactionbetween them. The term qcoupling can only be correctly omitted formaterials with a coherent intern structure. Even though models onthis coupling term are known [3], the term is neglected in mostliterature and research due to its complexity.

The thermal transport through a material according to thethermal gradient can then be quantified by the materials thermalconductivity ltot [W/(m K)]. A standard simplified approach forthis ltot is again to assume that the value represents a sum of single

ht] a comparison between a vacuum insulation panels and conventional insulation

Fig. 2. The thermal conductivity of air as a function of the air pressure and the

average pore diameter of the medium. Notice that the small pore size of aerogel and

fumed silica reduce the gaseous conductivity even at the atmospheric pressure of

1000 mbar (redrawn from [99]).

Fig. 3. Gaseous thermal conductivity of air (mW/(m K)) as function of the materials

characteristic pore size and the gaseous pressure at a temperature of 300 K. The

values have been retrieved from Eqs. (4) and (5) representing the Knudsen effect.

Compare with Fig. 26 for krypton.

R. Baetens et al. / Energy and Buildings 42 (2010) 147–172 149

values which describe on their own one of the previous mentionedways of thermal transfer, however they have to be consideredsimultaneously to be correct. This can be derived by rewritingEq. (1), dividing both sides of the equation with dT/dx [4]:

ltot ¼ lr þ lcd þ lg þ lcv þdx

dTqcoupling

� �(2)

where lr describes the radiation transfer between internal poresurfaces, lcd the solid conduction within the material skeleton, lg

the gas conduction within the material pores and where lcv

describes the air and moisture convection within the pores. Forbuilding insulation materials, all these parameters should beminimized to result in a low overall thermal conductivity ltot ofthe material in general.

1.1. The vacuum

The most effective reduction of the gas thermal conduction lg

appears in a theoretical perfect vacuum, as proposed by Sir JamesDewar (1892) [5]. Here, the lg achieves its limit value of ‘zero’.Although such a perfect vacuum is pure theoretically, a lowpressure has a positive influence on the gas conductivity.

A vacuum can be used to reduce the thermal conductivity ofmost traditional insulation materials, while the gaseous thermalconductivity lg of an evacuated material will not only be a functionof the applied pressure but also from the core materialscharacteristics. The gaseous thermal conductivity of a porousmedium at lower pressure is determined by the number of gasmolecules (determined by the particle frequency of the vacuum orthe internal pressure) as well as by the number of obstructions forthe gas on the way from the hot to the cold side. While reducing thegas pressure in a material, the gas conductivity of the non-convective gas remains almost unaffected until the mean free pathof the gas molecules reaches values in the same order of size as thelargest pores in the medium. When the pore diameter of thematerial becomes less than the average free length of path of gasmolecules, the air molecules will only collide with the poresurfaces without transferring energy by this elastic impact (seeFig. 2). Eq. (2) can than be reduced to the first two terms:

levac ¼ lr þ lcd (3)

However, to reduce the gas conductivity in conventional insulationmaterials as mineral wool, the pressure has to be reduceddrastically to the range of 0.1 mbar or below and the thermalconductivity will rapidly increase with increasing pressure.Therefore, a nano-structured core material in combination withthe pressure reduction is more favourable to be used in VIPs. Insuch material, a fine vacuum [6] is already adequate to reduce thegas thermal conductivity in a such a medium to a calculation valuelg = 0 W/(m K). Such a vacuum has a pressure around 10�3 bar, aparticle frequency of 1010 m�3 and a free length of path of 10�4 mand the thermal conductivity of the material will stay affected bythe reduced gas pressure up to one tenth of an atmosphere.

The influence of the air pressure on nano-porous materials canbe expressed analytical. The gas conduction in a porous media canbe written as follows [7]:

lg ¼lg;0

1þ 2bKn(4)

where

Kn ¼lmean

dand lmean ¼

kBTffiffiffi2p

pd2gPg

(5)

where Kn the Knudsen number is the ratio between the mean freepath lmean of air molecules and the characteristic size of pores d,

where dg is the diameter of the gas molecules and b a constantbetween 1.5 and 2.0 characterizing the efficiency of energy transferwhen gas molecules hit the solid structure of the material. Theconstant b depends on the gas type, the solid material and thetemperature. Because of the high porosity of insulation materials,the contribution of the gas conductivity lg will play an importantrole in the effective thermal conductivity at atmospheric pressure.However, the free air conduction lg,0 in Eq. (4) will be stronglyreduced due to the Knudsen effect if we consider the narrow poresize in nano-porous materials.

If we rewrite the previous Eq. (4), a formula which accentuatesthe three main parameters for gaseous heat conduction in porousmedia appears [8]: The gas pressure Pg, the characteristic pore sized and the temperature T (see also Fig. 3).

lc ¼ levac þlg;0ðTÞ

1þ CðT=dPgÞ¼ levac þ

lg;0ðTÞ1þ ðP1=2;g=PgÞ

(6)

where lg,0 is the thermal conductivity of free gas, P1/2,gas thepressure at which the thermal conductivity of the gas reaches thevalue of one half of lfree gas and C a factor defined as 2bkB=ð

ffiffiffi2p

pd2gÞ.

Equations for lg,0 and P1/2,g are given by Schwab [2]. The productPgd in Eq. (6) strongly defines the gas conduction: If micro- or nano-structured materials with small pore size are used as VIP cores,only a weak vacuum is required to reach a low thermal


conductivity. The value P1/2,gas in Eq. (6) is again stronglydepending on d and is influenced by the gas type.

1.2. The core

Because of the relationship between the gas thermal con-ductivity of air and the pore diameter, the core material has to fulfildifferent requirements to be suitable for vacuum insulation:

1. T

Figatm

fum

con

he materials pore diameter has to be very small. To reduce thegas conductivity in insulation materials with large pore sizes,the pressure has to be very low which is difficult to maintain byenvelopes primarily made of organic materials. Therefore, anano-structured core material in combination with a finevacuum is preferred in VIPs. Ideal would be a pore size of 10 nmor less, which would reduce the gaseous conductivity to zeroeven at atmospheric conditions.

2. T
he material has to have an 100% open cell structure to be able toevacuate any gas in the material.
Two other requirements can be found due to the specificcharacter of vacuum insulation panels:

3. T
he material has to be resistant to compression: Currentlyproduced VIPs have an internal pressure in the range of 0.2–3 mbar. Hereby, the pressure load on the panel is approximately1 bar or 100 kN/m2. The core material has to be stable enough sothe pores do not collapse when the panels are evacuated.
4. T
he material has to be as impermeable as possible to infraredradiation. This is necessary to reduce the radiation transfer inthe material to reach a very low conductivity value of the panel.
Many organic and inorganic insulation materials with an opencell structure are available to use as a core for VIPs. For each ofthem, a specific heat conductivity can be defined [99,9] as afunction of the gas pressure, as shown in Fig. 4.

Fig. 4 illustrates clearly that the use of conventional insulationas a core material for VIPs results in the necessity of a very highquality of vacuum (�0.1 mbar). Common organic envelopematerials cannot maintain this inner pressure for a long period:A rapid intake of air through the envelope will occur, resulting in afast increase of the thermal conductivity. Solutions to maintain thishigh quality of pressure almost always go together with anenvelope material with a higher thermal conductivity.

A material with very good achievement quality is pressedpowder boards made of fumed silica (as shown in Fig. 4), a fumedsilicon dioxide SiOx that is generally regarded for its unusualproperties. It has a low conductivity close to 0.003 W/(m K) up to

. 4. Thermal conductivity of different insulation materials as function of the

ospheric pressure (redrawn from [99]). Notice that the heat conductivity of

ed silica only rises above 50 mbar and that aerogel has a low thermal

ductivity at the atmospheric pressure of 1000 mbar.

50 mbar and has a conductivity of 0.020 W/(m K) at ambientpressure in dry conditions, half the thermal conductivity oftraditional insulation materials. The physical, hygric and thermalproperties of this material will be discussed.

1.2.1. Physical properties of fumed silica

The bulk density of a fumed silica material is in the range of160–220 kg/m3 which is nearly one order of magnitude higherthan the density of a traditional insulating material, but similar inweight as conventional insulation if the same total thermalresistance of a layer is assumed. Despite this, their porosity ishigher than 90% what means that the specific surface area is veryhigh: Commercial products have a specific surface area of 100–400 m2/g. However, measurements done by Morel et al. [10] haveshown that this specific surface area can decrease with values up to20% by ageing due to high relative humidity or high temperatures.

Important for VIP core materials are the pore size distributionPSD and the largest pore size diameter: These define the range ofvacuum necessary for the low thermal conductivity. Fumed silicamaterials have their largest pore size (300 nm) in the same order ofmagnitude as the mean free path of air molecules at standardtemperature and pressure (70 nm). Hereby, the material gasconductivity is even at atmospheric pressure affected by his finestructure.

The mean value of the specific heat of dry fumed silica is 850 J/(kg K), which is approximately the same value as a traditionalinsulation material like glass wool.

Powder boards of fumed silica have also a very low intrinsicpermeability k in the range of 2.6 � 10�5 to 3.0 � 10�5 m2/s at apressure gradient of 1 bar [2].

1.2.2. Water vapour adsorption of fumed silica

The adsorption-isotherm of fumed silica is derived [99] byfitting experimental results with an analytical model:

u ¼ 0:01721’0:08356þ ’

e2:82429’2:26663(7)

However, the sorption-isotherm of fumed silica is approximatelylinear with the relative humidity for a relative humidity RH up to50% (see Fig. 5). The slope of the sorption-isotherm can be

Fig. 5. The adsorption isotherm x ( pH2O) of fumed silica [upper] and the partial

water vapour pressure pH2O (T) of saturated air [bottom] if water vapour is allowed

to enter the core material (redrawn [98]).


approximated as defined in Eq. (8) within this range, as determinedby Schwab [2]:

uð’Þ ¼ a’ where a ¼ du

d’¼ 0:08 (8)

By considering the inverse function w(u) of the sorptionisotherm u(w) in Eq. (8), the internal water vapour pressure inthe material can be defined in function of the water content aspwv = w(u)pwv,sat(T) with pwv,sat the saturation pressure of watervapour at temperature T. The amount of absorbed water in fumedsilica powder boards will stay low (u < 0.05 kg/kg) for a relativehumidity lower than 60%. In this range, the water molecules willonly cover the surface of the silica grains by adsorption. However,for a higher humidity up to 95% will an exponential increase due tothe capillary condensation in the small pores be noticed. Animportant condition for building insulation is the moistureequilibrium at a relative humidity of 45%, the average relativehumidity of an indoor climate: A value u of approximately 0.04 kg/kg is found for fumed silica (see Fig. 5) which means that thegravimetric water content of the VIP core would never exceed thisvalue. This value matches approximately the saturation levelsbetween 0.03 and 0.07 kg/kg that are found by Schwab et al. [11] inapplied VIPs in typical German constructions (different valueswere found, depending on the structure and orientation).

This water content equilibrium, reacting on the changingoutdoor and indoor conditions, will affect the thermal conductivityof the core material: An increasing water content of the corematerial will result in an increase of the thermal conductivity ofthe VIP. The influence will be discussed in the next subsection,because it is important for estimating the service life of the VIPs.

1.2.3. Thermal conductivity of fumed silica

It has been said that the good thermal achievements of fumedsilica are caused to the structure of the material. The free airconduction lg,0 in Eq. (4) will be reduced due to the Knudsen effectif we consider the narrow maximum pore size in porous silica of300 nm: The mean free path of air lmean in normal conditions (23 8Cand 1 atm) is 70 nm which means that the Knudsen effect willstrongly reduce the gas conductivity in fumed silica withapproximately 40% at ambient pressure. Fumed silica has a typicalvalue P1/2 of �630 mbar while a conventional PUR foam has a P1/2

of 2.6 mbar, which proves again that fumed silica is very suitable asa core material in VIPs.

The dependency on temperature of the thermal conductivity iscommonly described as [12–14]:

levacðTÞ ¼ lcdðTÞ þ16

3

n2sT3r

EðTrÞ¼ lcdðTÞ þ lrðTrÞ (9)

where n is the index of refraction (�1 for low density fumed silica),s the Stefan–Boltzmann constant, lcd the conductivity of the solidmatrix, Tr the ‘‘Rosseland’’ average temperature within theinsulation material and where E [m�1] is the extinction coefficient.Tr [12] and E [m�1] [13,14] are described as:

T3r ¼ ðT1 þ T2Þ

T21 þ T2

2

4and ETr ¼ e�ðTÞreff ¼

reff K

rc

(10)

where T1 and T2 are the temperatures of the VIP surfaces and wherereff the effective density of the porous material, e*(T) the specificextinction coefficient as a temperature-dependent material valuecharacterizing the radiative attenuation. Furthermore, the extinc-tion coefficient E in Eq. (9) is the reciprocal of the mean free path oflength lph of thermal photons and has to be estimated by fittingexperimental data: For opacified silica cores, a value oflph � 100 mm is given by Fricke et al. [15] which means that VIPs

of 2 cm block off infrared radiation. More attention to the functionof opacifiers in fumed silica cores will be paid in Section 1.4.

Eq. (9) is introduced by Hottel and Sarofilm [16] and is appliedto VIPs by Brodt [4], where the second term of the equation wasdefined as the radiative heat conductivity lr. The equation is validfor all grey media, materials for which the mean free path forphotons is independent of their wavelength. It must be noticedthat also the thermal conductivity lcd of the core skeleton in Eq. (9)depends on the temperature. This dependency can be approxi-mated by lcd � Ta with a between 0.5 and 1.0 depending on thematerial. However, the influence of temperature on lcd will bemuch smaller than the effect on lr and will be mostly neglected. Anequation for this dependency is given [8], assuming that thedependency for fumed silica is similar as for silica glass:

lcdðTÞ ¼ 0:0021½ð�8:5� 10�12ÞT4 þ ð2:1� 10�8ÞT3 � ð1:95

� 10�5ÞT2 þ ð0:00883ÞT� (11)

Measurements [2,99] showed us that the thermal conductivity lcd

of the solid matrix of commercial fumed silica products in Eq. (9)comes out in the range of 0.0021–0.0034 W/(mK) and that theradiative heat conductivity can be found between 0.001 and0.004 W/(m K) at a low gas pressure of 1 mbar, depending on thetemperature.

Considering the application temperatures for building insula-tion, a linear approach can be made for the temperaturedependency of the thermal conductivity [17] taking into accountthe effect on the solid conduction and the effect on radiation:

levac;dry W=ðmKÞ ¼ ð0:0124T þ 0:0808Þ � 10�3 (12)

The dependency on water content of the thermal conductivity offumed silica is measured once [99,2,18] and shows a significantincrease of heat transfer through powder boards of fumed silicawith an increasing water content. For the specific staticcircumstances and panel sizes of the tests, an increase ofapproximately 0.5 mW/(m K) per mass percent of content isobserved in both papers. However, the value is corrected bySchwab [2] by deducting the increase of the gas conductivity due tothe water vapour pressure from the total thermal conductivityincrease to come to a lower limit of 0.29 mW/(m K) per masspercent water content. For the moisture equilibrium u of 0.04 kg/kg at ambient conditions of 50% relative humidity, a final thermalconductivity of 6 mW/(m K) can be found, starting with an initialthermal conductivity of 4 mW/(m K) for the dry core.

Because of the moisture increase in a VIP, three more types ofheat transport are possible: Heat conduction via water vapour witha partial pressure pwv, heat conduction by adsorbed water at theinner surface of the core and heat transfer via evaporation ofadsorbed water, diffusion and condensation of water vapour.Measurements [19,18] show us the results of the increase ofthermal conductivity per mass percent water content for a panelsize of 30 cm � 30 cm � 1 cm at a mean temperature of 10 8C, whatmakes it possible to define the thermal conductivity as function ofthe water content as in Eq. (13). However, the total effect ofmoisture on the thermal conductivity of a vacuum insulation panelis much more complex than the mentioned linear relationship.Complex non-linear relations between thermal conductivity,relative humidity, water vapour pressure and temperaturedetermine the effect of the moisture content on the overallthermal conductivity of VIPs [18,20,10,21].

@lc

@u� 0:05 W=ðmKÞ (13)

The total water content in a VIP will not be distributeduniformly in the VIP [18,20,22]: The partial water vapour pressurepwv will vary with the temperature gradient, conform the different

Fig. 6. Moisture distribution based on numerical simulations in a VIP with a

thickness of 20 mm and a total water content of 0.03 kg/kg (redrawn from [20]).

Note that the graph should be corrected including gravity.


pwv,sat for the corresponding temperatures. This distribution isnon-linear but can be approximated as linear for VIPs within a lowtemperature spread DT < 20 K. The spreading of the moisture isshown in Fig. 6 and the upper limit of the pwv (and as consequenceof lwv in Eq. (14)) is given by the coldest point in the VIP.

The consequences of this distribution can be clarified byexpanding Eq. (2) for dry VIPs with two more terms to include theeffect of moisture:

lvip;moist ¼ lcd;s þ lrad þ lg þ lhumidðuwðTðxÞÞÞ þ lwv (14)

where lhumid is the thermal conductivity by conduction ofadsorbed water in the core and lwv the thermal conductivity byheat conduction in water vapour. As shown in Fig. 6, lhumid willdepend on place due to the dependency of moisture content andthe spreading of it. As a consequence, the total thermalconductivity of a humid VIP should be written as different thermalconductivities in series or as a Riemann sum of l(x) divided by thepanel thickness.

However, the water gradient due to the temperature spread in aVIP will lead to a liquid transport back to the warm side. Asconsequence, the pressure gradient in the VIP should be inequilibrium with a vapour transport equalling the liquid transportunder stationary conditions. This rearrangement of moisture is alatent heat transport from the warm side of the VIP to the cold sideand introduces another type of heat transport in the VIP, called the‘heat pipe effect’.

The effect on the thermal conductivity can be expressed usingFourier’s second law expressed in Eq. (15) [18] describing the heatflow due to the temperature gradient and the enthalpy transport

Fig. 7. Cross sections of some typical envelope materials for VIPs, commonly named AF

metallized films) in scientific literature [106]. However, one must note that different t

due to the water vapour flow and using Eq. (16) describing the timeand position dependency of the moisture content due to aneffusion process driven by the potential r=

ffiffiffiTp

and a liquidtransport due to a water gradient:

rcc@T

@t¼ @lðuðxÞÞ

@x

@T

@tþ hD

@@x

DE1ffiffiffiTp @p

@x(15)

r@u

@t¼ @

@xDE

1ffiffiffiTp @ p

@tþ @

@xDK

@u

@x(16)

where hD is the enthalpy, DE the effusion transport coefficient andwhere DK is the liquid transport coefficient. The result of theequations can be seen in the corrected graph in Fig. 6 where theassumed water content distribution is corrected by the liquidtransport. Fitting both models of Eqs. (15) and (16) withexperimental data from Beck et al. [18] resulted in

DE ðkgffiffiffiffiKp

=ðms PaÞÞ ¼ 1:05� 10�11 and

DK ðkg=ðms PaÞÞ ¼ 4:03� 10�7 (17)

and a linear model for the dependency on the moisture contentdescribed as

lvip;moistðuÞ ¼ lvip;dry þ u ½kg=kg� � 0:0024 ½W=m K� (18)

It must be noticed that this thermal conductivity increment of0.24 � 10�3 W/(m2 K) per mass percent water content in the VIPretrieved in Beck et al. [18] approximates closely the correctedthermal conductivity increment of 0.29 � 10�3 W/m2 K per masspercent water of [2]. In addition, the heat pipe effect will increasethe heat transport, described by a pseudo thermal conductivitylHP. The same measurements [18] resulted in a lHP from 0.0005 upto 0.0017 W/(m K) for increasing moisture contents and increasingmean temperatures.

However, it is hard to predict the rate of increase of the totalwater content u: The increase depends on the water vapourpermeance of the foil, the climatic conditions and on the sorptionisotherm of the core material (defined in Eq. (8) and Fig. 7). So areby example different saturation levels of the core material between0.03 and 0.07 kg/kg found in typical German conditions [11],depending on the considered construction and orientation. Thismeans an increase of thermal conductivity between 0.0015 of0.0035 W/(m K) with Eq. (18) due to the intake of water vapour andalso means that the service life of a VIP cannot be definedunambiguous, but has to be expressed in function of theapplication conditions.

The same measurements [99] show a linear relation as inEq. (19) between the increase of the internal total gas pressure (seeFig. 4) and thermal conductivity in the range up to 100 mbar. Theprediction of the increase of internal air pressure and watercontent in building constructions [11] is based on the knowledge ofthe OTR and WVTR of the complete VIP envelope and on theatmospheric conditions as driving force. These envelope properties

(a metal film), MF1 (a single layer metallized film), MF2 and MF3 (both three-layer

ypes of foil with the same name are used in literature [98,24].


have been subject of research in [23] to come to a model of servicelife prediction and will discussed in Section 3.2.

The dependency on gas pressure of the thermal conductivity offumed silica is shown in Eq. (4) and the same measurements [99]show a linear relation between the increase of the internal gaspressure and thermal conductivity in the range up to 100 mbar:

@l@ pg

� 3:5� 10�7ðW=m K PaÞ (19)

This means a thermal conductivity increase of 0.001 W/(m K) for adry gas pressure increment of 30 mbar. Analytical models on botheffects (an increasing moisture content and an increasing inner gaspressure) and the influence of the effects on each other arediscussed in detail in Section 3 about the prediction of the VIPservice life.

1.2.4. Fire behaviour

Powder boards of fumed silica are not flammable M1 accordingto the French Standard NF P 92-510, which can be compared to the‘non flammable’ label A1 conform with the new Europeanclassification norms EN 13501-1 and the EN ISO 1182. Commercialavailable VIPs have a flammability label B2 according to DIN 4102,which is the lowest label of non-flammability that is acceptedwithin building applications.

1.3. The envelope

The outer envelope is one of the critical components of a VIP andis responsible to maintain the vacuum in the panel.

The envelope of VIPs is composed off multi layer films coveringthe whole element, including the edges. The use of an aluminiumlayer in these multi layer films is common because of the very lowgas and water vapour permeation. Due to the relatively highthermal conductivity of such an envelope, the heat flux increases atthe edges and corners. Because of this, the design of a VIP envelopewill not only be done from the point of view of gas and watervapour tightness, but also from the point of view to minimize thesethermal edge losses.

The multilayer films usable for VIP envelopes consist ofdifferent layers with an overall thickness of 100–200 mm.Currently, three different film types are being used for VIPenvelopes:

- M
etal foils consisting of a central aluminium barrier layer,laminated between an outer PET layer for scratch resistance andan inner PE sealing layer (see foil type ‘a’ in Fig. 7).
- M
etallized films made from up to three layers of aluminiumcoated PET films and an inner PE sealing layer (see foil types ‘b, c,d’ in Fig. 7).
- P
olymer films with different plastic layers laminated to each other.The gas and vapour permeation rate through these materials ishigher than with metal or metallized films. These films are onlyuseful if the required lifetime is not too extensive or if specialgetters are integrated in the VIPs (see also Section 1.4).
The permeability k [m2/(s Pa)] for air and moisture of theenvelope material is one of the determining criteria for the servicelife of the VIP: The pressure in the VIP should not rise above100 mbar after 30–50 years as a first rule. The permeationmechanisms of gases and water vapour trough a pure polymer filmdepend on the gas solubility coefficient S [m3(STP)/(m3 Pa�1)] andthe gas diffusion coefficient D [m2/s] of the foil material. Thepermeability is then given by their product k = DS [99]. The totalpermeability of a laminated foil can be calculated with thepermeability’s ki of the single layers of which the laminate consists

(analogue to Kirchhoffs Law for electrical conductivities in series):

1

ktot¼X 1

ki¼X 1

DiSi(20)

The permeability of air and the permeability of water vapourthrough a barrier layer depends on two different mechanisms: Thegas permeation predominantly occurs at macroscopic defects inthe range of 0.1–1 mm2, while the gas permeation through the bulkmaterial is practically not existing. In contrast to the dominatingfactor of macroscopic defects for gas permeation, the permeationof water vapour depends also on the microstructure of the layer:Capillaries can be formed at microscopic defects and grainboundaries.

To obtain a low permeable laminate by using the synergy effect,the two main technical requirements are

- a
low density of (microscopic) defects in the barrier layer and - a polymer possessing a low permeability next to the vacuum
coated layer.The barrier properties of a layer material are expressed in terms

of the water vapour transmission rate WVTR and the gastransmission rate GTR. Edge and corner effects have beenexamined on the WVTR and the GTR in VIPs [99,23], while alsodetoriation by heat and moisture loads at defects has been subjectof research [24,25]. The characteristics of these WVTR and GTR willbe further discussed in Section 3.2 about gas pressure and moisturecontent increment in VIPs.

A typical VIP envelope foil with three metallized layers has a airtransmission rate ATR of 2–5 � 10�10 m3(STP)/(m2 day) at 23 8Cand a 50% relative humidity and a water vapour transmission rateWVTR of 1–5 � 10�6 kg/(m2 day) at 23 8C and 85% relativehumidity [25]. However, measurement standards do not yet existfor the GTR and WVTR of vacuum insulation foils because it wasimpossible to measure these extremely low ranges of transmissionuntil in [99]. Moreover, fast measurement methods for very lowgas transmissions are proposed in [26] based on the heliumtransmission of the foils.

The influence of the choice of envelope material on the servicelife and on the thermal bridge effects will be discussed in Sections 2and 3.

1.4. Getters, desiccants and opacifiers

Important for the service life of the VIPs is maintaining the innervacuum. To increase their service life, getters and desiccants areoften added in the VIPs: Continuously adsorbing the gasses(getters) and the water vapour (desiccants) in the VIP corematerial, they prevent the increase of the internal gas and vapourpressure. The inner water vapour pressure and gas pressure willstay equal to the manufacturing conditions until the capacity of thegetters and desiccants is reached. As a consequence, they preventthe increase of the thermal conductivity due to the higher pressureand they increase the lifetime of the VIP. Some core materials ofVIPs have the possibility to fulfil the function of getters anddesiccants themselves, but not all of them. This makes it importantto add these chemicals to the core although they decrease thethermal resistance of the VIP and increase the manufacturing costs.

Opacifiers are added to the fumed silica in order to make itopaque to infrared and hereby to reduce the radiative conductivityto a low level. A common opacifier for fumed silica cores is siliconcarbide powder.

2. Thermal bridges

It is clear that VIPs cannot be seen as a material but that theyhave to be seen as a system of materials and properties. These have


all their influence on the total thermal performance and thepossible applications of vacuum insulation panels. A lot of researchhas been done on reducing and estimating the thermal bridgingand on increasing and estimating the service life time.

The better the insulation material the higher the importance ofthe heat flux due to thermal bridges. In a assembly insulated withvacuum insulation panels, we can notice three levels of thermalbridges: The level of the VIP due to the barrier envelope, the level ofthe building component (in VISs) and structural thermal bridges.The thermal bridge due to the envelope has been yet subject ofmany researches, while the structural thermal bridges with VIPinsulated details has not yet been subject of large scale studies. Thefirst two types of thermal bridges will be discussed in detail in thenext two sections.

2.1. Thermal bridge effect on the scale of VIPs

Thermal bridges on the VIP level are due to the continuing of thethin high barrier envelope from the cold to the warm side of the VIPand due to small air gaps between two adjacent panels. Both have anot unimportant influence on the overall thermal performance of aVIP insulation layer: Due to this thermal bridge, the effectiveconductance Leff of the vacuum insulation panel is higher than thecentre-of-panel thermal conductance that is determined in Section1.2.1:

Leff ¼Lcop þcedge

P

A(21)

The linear transmittance cedge in this equation depends onthe panel thickness d, the centre-of-panel thermal conductivitylcop, the barrier film thickness tf and the equivalent foilthermal conductivity lf. This will result in different valuesfor cVIP depending on the laminate layers and their thermalproperties.

An analogue expression can be written with the direct heattransmission coefficient Hdirect of an assembly [27,28] or insulationlayer according ISO 14683:2007(E):

HD ¼X

i

UiAþX

k

ck;lineli þX

j

x j;pointn j

0@

1A (22)

The value Ui of a wall without the thermal bridges, can bewritten as:

Ui ¼1

ainþX

l

dl

llþ 1

aout

( )�1

¼ 1

ainþX

l

1

Llþ 1

aout

( )�1

(23)

where ai is the heat transmission coefficient at boundary surface i.This show us that the influence of the thermal bridge depends onthe composition of the other layers of the wall and on the heattransition coefficients at the surface. Notice that cline in Eq. (22)will be zero for the VIP edge if Ui is calculated with the Leff fromEq. (21).

cvip;edge;0 ¼lcffiffiffiffiffiffiffiffiffiffiffia1dp

p þ 1þ lcffiffiffiffiffiffiffiffiffiffiffia2dp

p( )�1

:a1ðN2

2 � BÞ

ð’dplf=l0f ÞðN2

1N22 � B2Þ � l1

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiN2

1N22 � B2

qðð2B=

ffiffiffiffiDpÞ � 1Þ � l2

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiN2

1N22 � B2

qð1� ð2B=

ffiffiffiffiDpÞÞ

264

375 (25)

2.1.1. Thermal bridging due to VIP envelope

Measurements [100,27] show values of the linear transmit-tance cedge from 0.001 to 0.400 W/(m K) and an effectiveconductivity leff in the range of 0.0051–0.0086 W/(m K) for VIPswith a panel size of 1.00 m by 0.50 m by 0.02 m and with a centre-

of-panel conductivity lcop of 0.004 W/(m K), depending on theproperties of the envelope material. This wide range of the datashow us the importance of the prediction of the influence of thepanel edges to come to a calculation value of the overall thermalconductivity of an insulation layer or assembly. Analytical models[28–30] and numerical investigations [28,31–34] have beenproposed on the calculation and prediction of this thermal bridgeeffect. Numerical investigations or parts catalogues are more likelyto be used to be used to estimate thermal bridge effects in thecommon application of VIPs [35–40], while analytical models areof more interest to be programmed in an ordinary spreadsheets.

2.1.1.1. Analytical models. Originally, the high barrier laminateconsisted of stainless steel or glass [41] envelopes, aluminium foilsor laminated polymer films but gradually developed intolaminated metallized polymer films with a low GTR and WVTRto decrease the thermal conductivity of the foil. The analyticalmodels [29,30] consider the thermal bridges due to the VIPenvelope and they are developed to understand the systems of heattransfer at the panel edges, which is impossible if only numericalsimulations are used.

The models of [29,71] start with the same formulae of the directheat transmission coefficient as in Eq. (22) as Schwab et al. [27] andISO 14683:2007(E). The third term of the equation with the pointthermal transmittance is neglected in the model, assuming that thecorner thermal bridge effect is much smaller than the effect of thepanel edge. The equation clearly shows us the importance of thesize and shape of the VIP panels: The larger the panels and thesmaller the perimeter length, the smaller is the influence of thelinear thermal conductivity of the barrier envelope on the overallU-value of the insulation layer. In many older texts, a minimumsize of 0.50 m by 0.50 m for VIPs is proposed. Two single stepanalytical models for estimating the cvip,edge are found [29].

- T
he simplified model of [29,71] starts from the assumption thatthe thermal conductivity of the core material lc equals 0.0 W/(m K), which is valid as long as lc or the ratio lc/lf is sufficientlysmall. This equals the assumption that the energy flux throughthe bulk material is zero and hereby that every energy flux at theedge is caused by the thermal bridge of the barrier envelope. Theequation for the linear transmittance is then given by
cvip;edge;0 ¼1ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

a1dflf

p þ ’dp

dflfþ 1ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

a2dflf

p( )�1

(24)

where an is the heat transmission coefficient at boundary surface

n, dp the thickness of the VIP, df the thickness of the laminate, d0fthe thickness of the laminate at the panel edge, w the ratio of

df=d0f , and lf the equivalent laminate thermal conductivity of the

VIP laminate.
- T he advanced model of [29,71] is given to make a general equation
for cvip,edge for models for which the previous assumption in thesimplified model cannot be made. The equation for the linearthermal transmittance of the VIP envelope is here formulated as:

where l0f is the thermal conductivity of the laminate at the paneledge. The result of this equation is graphically shown in Fig. 8. Adistinction is made between the thermal conductivity of theenvelope foil and the VIP edge to include the possibility that theedges are reinforced, e.g. with a tape. In Eq. (25) are Ni and B

Fig. 8. Behaviour of the linear thermal transmittance as function of the panel

thickness df for different VIP envelopes and for a centre-of-panel thermal

conductivity of 0.004 W/(m K) (redrawn from [61]).

Fig. 9. Behaviour of the linear transmittance as function of the panel thickness for

different centre-of-panel thermal conductivities and a 6 mm thick aluminium

barrier (redrawn from [61]).


parameters which are defined as:

Ni ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiai

dflfþ lc

dflf dp

sand B ¼ lc

dflf dp(26)

and where l1 and l2 are eigenvalues of the linear system of

differential equations derived [29] to represent the thermal

phenomenon. Two remarks can be given on these Eqs. (24) and

(25): If the limit for Eq. (25) is taken for lc! 0, the Eq. (24) for

zero thermal conductivity of the core material appears, and both

Eqs. (24) and (25) derive the model of the linear thermal

transmittance of the edge of one single panel. If two panels are

adjoined, the sum of the linear thermal transmittances of both

panels have to be taken, to come to the total linear thermal

transmittance of the joint.
- T he parameters which can influence cvip,edge can now be
discussed. The linear thermal transmittance of the VIP envelopedepends on four parameters: The laminate thicknesses df and d0f ,the laminate thermal conductivity lf, the core material thermalconductivity lc and the panel thickness dp.

The influence of the laminate material on the cvip,edge is dual:The laminate thickness df has an influence and the laminatethermal conductivity lf has his influence (see Fig. 8). First, thelinear thermal transmittance cvip,edge increases for an increasinglaminate thickness df: An increase in laminate thickness from 6 to20 mm can increase the cvip,edge with 130% [29]. Secondly, thelinear thermal transmittance cvip,edge increases for an increasinglaminate thermal conductivity lf.

The influence of the core material is unambiguous in Eq. (25):An increasing thermal conductivity of the core material lc makesthe linear thermal transmittance decrease (see Fig. 9). The effectof this is stronger as the panel thickness is smaller. The heat fluxthrough the bulk material of the VIP will become larger, whichmakes the relative importance of the heat flux through the paneledge decrease.

The influence of the seam modelling can be calculated with theanalytical model of [29,71] with the right choice for the ratiow. Theratio ’ ¼ df=d0f can be used to integrate the thickness difference inthe laminate at the seam, but a limitation to this thickness d0f of thepanel edge has to be introduced: If this thickness increases, anadditional heat flow over the width of the foil has to be considered.However, this limitation will disappear when using the analyticalmodel for building components (see also Section 2.2).

- T
he limits of the analytical models are checked by comparingmodel predictions with numerical data: A deviation <5% foridealized envelopes and <7% for realistic seams has been found.This proves that the model is a good alternative for complexnumerical models to predict the linear thermal conductivity ofthe barrier envelope in film based VIPs with an envelope andedge thickness of 10–300 mm. However, significantly largerdeviations were noticed for VIPs with a barrier envelope of two-or three-layer metallized films (i.e. whereas three-layer metal-lized films are recommended in [99]), a centre-of-panel thermalconductivity of 0.020 W/(m K) and a panel thickness less than20 mm.
2.1.1.2. Numerical simulations. Numerical simulations have beenmade to predict the linear thermal conductivity of the VIP envelope[42,27,29,32–34,65]. Here, the overall heat flow wq,total [W] iscalculated using thermal analysis software and the correspondinglinear thermal transmittance is obtained from

cvip;edge ¼fq;tot � fq;cop

l2D DT¼

fq;tot

l2D DT� Ucopbp (27)

where wq,cop [W] is the centre-of-panel heat flow, DT thetemperature difference, bp the simulated panel width and wherel2D is the simulated length. The simulated panel width is 200 mm inolder articles, but has been shown [42] that this was not adequateto obtain adiabatic boundary conditions for AF-laminates whereasa simulated panel width of 500 mm is [29,34].

The similarity between the results of the analytical models andthe results retrieved from numerical simulations with widespreadcomputer programs was used to prove the accuracy of theanalytical model, whereas the numerical simulations have beenvalidated previously [43].

2.1.1.3. Discussion. The result of both numerical and analyticalmodels show us that laminated aluminium foils (foil type AF inFig. 7) have a big influence on the total overall conductance of theVIP (see Table 1): Values up to 50 times higher are found for cedge

for AF-VIPs compared to aluminium coated multilayer foils if no airgaps between the panels are assumed (see Section 2.1.2). Theincrease of the overall thermal conductance will range values of360% for this foil, while values between 2 and 44% are found for theAl coated multilayer foils (foils MF1, MF2 and MF3 in Fig. 7).

Table 1Thermal and linear thermal conductivities [W/(m K)] of the four main foil types for vacuum insulation panels in function of the panel thickness [mm]. The values are retrieved

from equation Eq. (24) and slightly higher compared to the values retrieved from [100] due to the neglection of heat transfer through the core.

lf 5 mm 10 mm 15 mm 20 mm 25 mm 30 mm 35 mm 40 mm 45 mm

AF-VIP 25 0.0760 0.0660 0.0583 0.0522 0.0473 0.0432 0.0397 0.0368 0.0343

MF1-VIP 0.38 0.0045 0.0028 0.0021 0.0016 0.0013 0.0011 0.0010 0.0009 0.0008

MF2-VIP 0.42 0.0049 0.0031 0.0023 0.0018 0.0015 0.0012 0.0011 0.0010 0.0009

MF3-VIP 0.90 0.0087 0.0059 0.0044 0.0036 0.0030 0.0025 0.0022 0.0020 0.0018


Improvements on this seam edge have been proposed in severalpapers: Numerical models show us [27,31] that the high linearthermal conductivity of the edge of these AF-VIPs can be reduceddrastically by encapsulating the VIP with an other insulationmaterial such as expanded polystyrene (EPS) or extrudedpolystyrene (XPS). However, the obtained values will still exceedthe linear thermal conductivities of VIPs with aluminium coatedmultilayer foils. For these laminates, another – so far theoretical –improvement is the serpentine edge [44,31] where the path for theheat flux is prolonged with a reduced heat flow through the edge asa result. It was proven that by optimizing the depth and increasingthe amount of slots, the linear thermal conductivity of the edge canbe decreased drastically. However, the practical consequencessuch as feasibility and thermo-hygric behaviour of this edge modelare very complex and have not been studied yet.

2.1.2. Thermal bridging due to air gaps between two adjacent

envelopes

Because of the irregular shapes of the VIP edge, it is possible thatan air gap occurs between two adjacent VIPs. The influence of thisgap was investigated [28,27] and no air leakage from one side ofthe panel to the other side through the gaps was assumed. Itbecame clear that the influence of the air gap with laminatedaluminium foils has a minor influence on the linear thermaltransmittance for laminated aluminium foils: An increase in therange of 7–15% of cvip,edge is noticed for air gaps up to 5 mm.However, an average value of cvip,edge = 0.320 and 0.170 W/(m K)is noticed for panel-thicknesses of respectively 10 and 20 mm.Therefore, VIPs with laminated aluminium foils should not be used

cðcÞedge ¼a1

ðT1 � T2Þ� wðT1 � TsxÞ �

B1ðTsy � c0yÞðl1 � l2Þ þ B1B2ðTsx � coxÞððl2=ðN22 � l2

2ÞÞ � ðl1=ðN22 � l2

1ÞÞÞffiffiffiffiffiffiffiCDp

" #(28)

in buildings if the panel size is smaller than 1 m2 due to the highrelative importance of the linear transmittance of the edges.

Compared to the aluminium foil laminates, the cvip,edge

changes more with the variation of the air gap when alumi-nium-coated multilayer foils are used: An increase in the range of600–900% of cvip,edge is noticed for air gaps of 5 mm compared to0 mm air gaps. However, the value of the cvip,edge stays in the rangeof 0.012–0.022 W/(m K) which is a factor 10 smaller compared tolaminated aluminium foils. Hereby, the gap for aluminium coatedmultilayer high barrier foils should be as small as possible to makeproper use of the high thermal resistance of VIPs.

2.2. Thermal bridge effect on the scale of a building component

A building component is defined as a VIP fixed in two protectionskins linked with a spacer. This spacer is a link between the innerprotection skin and the outer skin, which is a clear example of athermal bridge. Numerical simulations have been done[42,31,32,45] and show us what one could expect: The lowestlinear thermal conductivity cspacer can be found with facings andspacers which have the lowest thermal conductivity and the

cspacer is in all cases higher than the calculated cedge at the VIPlevel.

The best results are found for a plastic tape as a spacer (seeFig. 10). However, this tape is no structural link between the twofacings: Loads are transferred by gluing the facings and the VIPtogether and the tape will only have a safety function in case theglue fails due to ageing or in case of mechanical damage.Depending on the size of the panels and the face materials, valuesfor cspacer in the range of 0.001–0.005 W/(m K) can be found[42,46] for VIP panels with a thickness of 20 mm due to the plastictape as a spacer.

If the protection skins on one or both sides of the VIP have a highthermal conductivity a ‘‘heat drainage effect’’ will occur along thesurface towards the spacers. Result of this is the very high values ofcspacer for components with facings in glass or steel: Values in therange of 0.010 up to 0.020 W/(m K) [32] occur for metal facings.

An extension of the model of [29,71] about thermal bridges dueto the high barrier envelope has been proposed to include the effectof the facing materials and the edge spacer: A heat flow over thesurface of the panel edge is considered. This model is discussed inthe next subsection.

2.2.1. Analytical model of Tenpierik et al. [30]

Two modifications are made [30] on the previous model to meetto the scale of a building component: First, the heat balance at thecorners of the component is modified. Second, the equation for thelinear transmittance is extended to include the additional heatflow through the edges. The linear thermal transmittance of thethermal bridge is now calculated as:

where w is the width of the panel edge and where Ni and Bi areparameters defined as

Ni ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ai

df ;ilf ;iþ lc

df;ilf ;idp;i

sand Bi ¼

lc

df ;ilf ;idp;i(29)

and where l1 and l2 are the eigenvalues of the linear system of thedifferential equation derived to represent the thermal phenom-enon [29]

l1 ¼ �

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiðN2

1 þ N22Þ �

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiðN2

1 � N22Þ þ 4B1B2

q2

vuut;

l2 ¼ �

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiðN2

1 þ N22Þ þ

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiðN2

1 � N22Þ þ 4B1B2

q2

vuut(30)

and with D the discriminator of the second square root of theeigenvalues and C a parameter:

D ¼ ðN21 � N2

2Þ þ 4B1B2 and C ¼ N21N2

2 � B1B2 (31)

Fig. 10. Four edge spacer construction types: (a) aluminium spacer of double glazing, (b) folded edge construction, (c) thermoplastic spacer and (d) reinforced non-metallic

tape [30]. A similar division is made in [98] and Bundi [107].


The terms c0x and c0y are temperatures of the face sheets at a placewhere only the one-dimensional effect of the VIP occurs and Tsx

and Tsy are fictive temperatures of the face sheets in front of thethermal bridge. The resistance of the edge spacer is included in theequation to determine the fictive temperatures of the face sheets infront of the thermal bridge and the analytical formulae for thosevalues are given by Tenpierik et al. [30]. It seems that the linearthermal transmittance depends on the environmental tempera-tures, but it can be proven that, after some rewriting, that these areeliminated from the equation.

Example studies on thermal bridges in high performancebuilding components have been investigated for four differentedge spacers [42,33,46,32,30] in function of facing materials andbarrier envelopes as shown in Fig. 10. Remarkable in the analysis ofthese four examples is the increasing thermal conductivity of thealuminium spacer for an increasing panel thickness from 3.75 W/(m K) for a 10 mm VIP to 9.61 W/(m K) for a 40 mm VIP, while thisvalue for the three other edge types stays constant within in thesame range of panel thickness. This means that the maximumvalue for the linear transmittance according to Fig. 8 can be foundat thicker panels for an aluminium spacer. However, for the threeother edge spacers are the calculated cspacer a factor 10 smaller andas result more interesting for applications from which can beconcluded that aluminium spacers are less suitable for with VIPinsulated facade panels. The best results are found for a reinforcednon-metallic tape with a ltape = 0.33 W/(m2 K) and dta-

pe = 0.15 mm as an edge spacer (edge spacer ‘c’ in Fig. 10).

- T
he limits of the analytical models are checked by comparingmodel predictions with numerical data. The deviations betweenboth data were smaller than 5% for building components with afacing with high thermal conductivity. However, the deviationsincrease for low conductivity facings but remain below 10%,caused by errors in the approximation of the edge constructionand simplifications in the equation: The edge construction has tobe characterized by two parameters (the equivalent edge thermalconductance K and the edge width) to use the equation. It is in thecalculation of this equivalent thermal conductance K that aschematization error occurs. Hereby is the analytical model errorsignificantly smaller than the total noticed deviation. Besidesthis, the deviations on the linear thermal transmittance will staylower than 5%, except if the linear thermal transmittance is lowerthan 0.1 W/(m K).
The error analysis [30] indicates that the possibility to use theequation strongly depends on the ratio j, the product of foilthermal conductivity and thickness to the product of facingthermal conductivity and thickness and on the equivalent edgethermal conductance K.

2.2.2. Numerical models

Analogue as for the VIP edge is the similarity between theresults of the analytical models and the results retrieved fromnumerical simulations used to prove the accuracy of the analyticalmodel.


2.3. Thermal bridge effect on the scale of the building

The relative importance of structural thermal bridges on theoverall thermal performance of assemblies increases strongly withthe quality of the insulation material. Compared to the thermalbridges of the other levels, the structural thermal bridge has thelargest influence on the overall thermal properties of an assembly.Building details have been sporadically studied based on numericalsimulations [27,32–34] or in a large-scale study [45,43].

The small studies on the building details show the highpotential of vacuum insulation panels but show also that joints andconnections with other components have to be solved carefully:The risk on condensation damage is higher due to possible lowsurface temperatures and air leakage. The thermal bridges will alsodepend on the edge effect on the vacuum insulation panels, whichmeans that the edge effect of the VIP should have to be consideredin combination with (every possible) other material(s) that makecontact with the edge. However the thermal behaviour ofstructural thermal bridges is already complicated, also thecombined thermo-hygric behaviour should be studied for thisdetails. Due to the complication of this models, the best thermalsolutions should be looked for first and to continue afterwardswith the combined thermo-hygric behaviour of the details.

3. Service life prediction for VIPs

The service life of vacuum insulation panels depends on severalfactors. A first important factor is the assumed definition of theservice life and the assumed requirements according to thisdefinition. Secondly, the core material will have a great influenceon the service life, determining the increase of the thermalconductivity of the VIP according to inner gas pressure and watercontent. A last important factor is the type and the quality of theVIP envelope around the core and the atmospheric conditions inwhich the VIP is applied, which will determine the increase of theinner gas pressure and water content.

3.1. Service life definition

Two different definitions for the service life tSL can be found inliterature (see also Fig. 11). The first and most commonly useddefinition – e.g. by manufacturers – for the VIP service life can bedescribed as ‘the time elapsed from the moment of manufacturingt0 until the moment the effective thermal conductivity leff of the

Fig. 11. Service life definition of a vacuum insulation panel (redrawn from [108]).

Note that the slope of the increasing thermal conductivity is not realistic compared

to the real slope.

vacuum insulation panel has exceeded a certain limiting value llim

(t = tSL)’ [101,47] (ASTM C1484-01) or

leff jt¼tSL¼ llim (32)

The second – less used – definition can be described as ‘the timeelapsed from the moment of manufacturing t0 until the momentthe time-averaged effective thermal conductivity of the vacuuminsulation panel equals a critical value lcritical (t = tSL)’ or

leff

��t¼tSL

¼ lcritical where leff ¼1

t

Zt0

leffðtÞdt (33)

The thermal conductivity limit is in general defined for the corematerial lc, to exclude the effect of the thermal bridge. This limitthermal conductivity is generally defined as 0.008 W/(m K) [48] oras 0.011 W/(m K) in the American standard ASTM C1484-01,starting from an in general accepted centre-of-panel thermalconductivity of 0.004 W/(m K) for a dry VIP. A referencetemperature T0 of 296 K should be used for the service life becausethe regression properties of the core material are defined at thistemperature.

Different ageing mechanisms can be assumed to predict thisservice life of vacuum insulation panels. A first failure of the VIPcan occur in the first days after manufacturing: The seams of theVIP can be established to be of low quality with a rapid increase ofthe internal gas pressure as consequence or the envelope canbecome damaged during the installation in the assembly or on site.The chances on the first type of failure will be reduced bypreserving the VIPs the first 10 days in the factory, the secondpossibility of failure can be expelled by pre-fabricate theassemblies in a factory. However, both failures are neglected inthe prediction of service life, assuming that the damaged VIPs willbe excluded before application.

During service, pressure increase will occur due to slowpermeation of gas molecules through the envelope. Also will thewater content of the VIP core increase due to intake of moisturethrough the barrier. Both effects will be postponed or slowed downby the added getters and desiccants in the core, but still they willresult in an increase of the thermal conductivity of the VIP core.These three factors (the pressure increase, the increase of watercontent and the influence of getters and desiccants) are thecommon factors used to define the service life of a VIP for a specificlambda-criterion. The change in thermal conductivity can than bewritten for constant environmental conditions as Eq. (34) [49]which is an extension of the equation of Brunner and Simmler[52,53].

dlc

dt¼ @lc

@ pg

d pgðT;’Þdt

þ @lc

@ pwv

d pwvðT;’Þdt

þ @lc

@u

duðT;’Þdt

(34)

Every term of these equation can be studied in detail to come to afinal equation for the service life of a vacuum insulation panel.

3.2. Thermal conductivity of the core as function of moisture content

and air pressure

In Section 1.2.1, we have seen that the thermal conductivity ofthe core material depends on the air and moisture pressure in thematerial. A first approximation of the dependency was expressedin Eqs. (13) and (19). Later on, the dependency has beeninvestigated in detail [50,17,19]. The conclusions on the pressureand moisture dependency of the thermal conductivity can besummarized by Fig. 12.

Fig. 12. Thermal conductivity (mW/(m K)) of the common core material fumed

silica as function of both the air pressure (mbar) with an external load of 1 bar on

the sample and the water content (mass%), at a mean temperature of 20 8C.

Fig. 14. Calculated increase of the total air pressure in VIPs defined using Eqs. (37),

(46) and (47), for panel sizes of 50 by 50 by 1 cm3 and 100 by 100 by 2 cm3 and for

three different foil types AF, MF1 and MF2 (see Fig. 7). The graph is calculated with

values for the envelope GTRL and GTRA determined at 23 8C and 75% relative

humidity, i.e. 0.0018 (AF), 0.0080 (MF1), 0.0039 (MF2) cm3/(m day) and 0.016 (MF1)

cm3/(m2day) respectively [51] (GTRA has been found negligible for AF and MF2, i.e.

assumed zero in the calculations), and with a porosity of 90% for the fumed silica

core. The inner air pressure is assumed to be zero at t0 = 0 years whereas

commercial products have an inner air pressure between 0.5 and 5 mbar and no

getters and desiccants have been taken into account. One assumes that the VIP

envelope properties remain the same, i.e. that no degradation will occur through

time, which is unlikely (see Section 3.5).


3.3. Pressure and moisture content increase as function of envelope

material properties

The total gas transmission rate GTRtot [m3(STP)/day] determin-ing the amount of gas that permeates into the VIP per given time isdefined from the measured total permeance Qgas [m3(STP)/(m2 day Pa)] of the envelope foil [23]:

GTRtot ¼ ðAQg;A þ LQg;L þ Qg;PÞD pg;e

¼ GTRAAþ GTRLLþ GTRP (35)

where subscript P denotes point defects. A relation between thepressure increase dpg/dt and the total permeance Qg,tot or theGTRtot can be written from the law of mass conservation and with

Fig. 13. Calculated increase of the total water content in VIPs defined using Eqs. (40),

(46) and (47), for panel sizes of 50 by 50 by 1 cm3 and 100 by 100 by 2 cm3 and for

three different foil types AF, MF1 and MF2 (see Fig. 7). The graph is calculated with

values for the envelope WVTRA determined at 23 8C and 75% relative humidity, i.e.

0.0006 (AF), 0.035 (MF1) and 0.0086 (MF2) g/(m2day) respectively [51], a dry core

density of 200 kg/m3, u(w) as defined in Eq. (8) and psat as defined in Fig. 5, and

matches the short-term increase of water content as defined by Schwab et al. [51].

No getters and desiccants have been taken into account and one assumes that the

VIP envelope properties remain the same, i.e. that no degradation will occur

through time, which is unlikely (see Section 3.5).

the ideal gas equation (see also Fig. 13):

d pg

dt¼

Qg;tot D pg;e

Veff

Tm p0

T0

� �¼ GTRtot

Veff

Tm p0

T0

� �(36)

Solving the differential equation in Eq. (36) for pg yields:

pg ¼ pg;e � ð pg;e � pg;t¼0Þeð�Qg;tot=Veff ÞðTm p0=T0Þt (37)

where Dpg,e = pg,e � pg is the difference between the atmosphericpressure and the inner VIP gas pressure, and Veff the total porevolume, which can be obtained by multiplying the core volumewith its porosity and where (Tmp0)/T0 is a conversion factor fromstandard conditions to measurements conditions. The result of thisequation is graphically shown in Fig. 14.

An analogue relation can be written for the water vapourtransmission rate WVTR [g/day] as:

WVTRtot ¼ Qwv;tot D pwv ¼dXw

dt(38)

where Dpwv is the difference between the inner VIP water vapourpressure. The penetrated water vapour will be partially absorbed bythe fumed silica. The dXw/dt will depend on the adsorption of wateror water vapour of the core material. This adsorption is expressed inEq. (8) and expresses as a consequence also the relation betweenQwv,tot and Dpwv of Eq. (38). The term dXw/dt can then be written as:

dXw

dt¼

Qwv;tot

mvip;drypwv;satðTÞ½’wv;out � ’wv;inðXÞ� (39)

where mvip,dry is the dry mass of the VIP core and wwv,out and wwv,in

are the relative humidity inside and outside the VIP. This equationhas an analytical solution Xw(t) [51] assuming the linear relationXw = kw of Eq. (8):

XwðtÞ ¼ k’wv;outð1� eð�Qwv;tot pwv;satðTÞ=mvip;drykÞtÞ (40)

The result of this equation is graphically shown in Fig. 13for a core material with a saturated water content of 6%. It

Table 2Influence of the relative humlidity on the WVTR of a VIP envelope (Schwab, 2005c).

Panel size Conditions Foil AF Foil MF1 Foil MF2

WVTRA [g/(m2 day)] 10 cm�10 cm�1 cm 25 8C, 45% RH 1�10�3 9.6�10�3 1.5�10�3

25 8C, 75% RH 2�10�3 36�10�3 18�10�3

Table 3Influence of the temperature on the GTR of a VIP envelope (Schwab, 2005c).

Panel size Conditions Foil AF Foil MF1 Foil MF2

GTRtot/L [m3/(m day)] 20 cm�20 cm�1 cm 23 8C, 14 mbar 3�10�9 9�10�9 1.5�10�9

45 8C, 14 mbar 7�10�9 36�10�9 2�10�9

65 8C, 14 mbar 10�10�9 49�10�9 4�10�9


becomes clear that water content will not infinitely increase butwill evolve to the saturated water content, while the inner airpressure will increase approximately linearly with the time in thefirst decades.

The values GTR and WVTR are foil properties and can beobtained from measurements done [23] for the laminatedaluminium foil AF and the aluminium coated multilayer foilsMF1 and MF2 (see Fig. 7). However, the results of thesemeasurements depend significantly on temperature, relativehumidity and panel size.

3.3.1. Dependency of WVTR and GTR on panel size

As seen before can we write the ATRtotal as function of the size(area A and perimeter L) or the side K for a square panel as Eq. (35)or as

ATRtot

L¼ ATRA

K

2þ ATRL (41)

No dependency of ATRtot/L on the panel size is found in [23] for theAF and MF2 foil types (see Fig. 7), from which can be concluded thatinfusion of gasses mainly occurs through the VIP edges. For MF1-covered VIPs, the ratio ATRtotal/L increases with an increasing panelsize, which means that surface permeation contributes to the totalinfusion. In general, it may be concluded that the gas permeationwill mainly occur through the seams of the VIP specified by ATRL,due to the dependency of gas permeance on the macroscopicdefects as mentioned previously in Section 1.3, while area relateddefects will be of at least same order of importance for large panels,e.g. above 1 m2 [99].

This conclusion is not valid for the WVTR: Water vapourpermeation will occur through the barrier layer as well as throughthe seams due to the dependency of microscopic defects. However,it is very difficult to separate the water vapour permeation acrossthe surface and the edge of a vacuum insulation panel. Only anupper value of the WVTRA and the mass increase can be estimatedif the length related permeation is neglected:

WVTRA �WVTRtot

A(42)

Table 4Properties of high barrier foils AF, MF1, MF2 and MF3 [25] (Schwab, 2005c). It must be not

of 50%.

Measurement conditions

T pH2O Size [cm3]

GTRA [m3/(m2 day)] 25 8C 14 mbar 20�20�1

GTRL [m3/(m day)] 25 8C 14 mbar 20�20�1

GTRL [m3(STP)/(m day)] 25 8C 14 mbar 20�20�1

Pressure increase [mbar/year] 25 8C 14 mbar 20�20�1

100�100�2

Ea [kJ/mol]

WVTRA [g/(m2 day)] 25 8C 14 mbar 20�20�1

An example of the influence of the panel size on the water vapourpermeance can be noticed in Fig. 14 where the increase of watercontent is expressed for 2 different panel sizes.

3.3.2. Dependency of WVTR on relative humidity

A MF1 foil has a typical WVTRA of 0.0096 g/(m2 day) [23] at25 8C and at the ambient water vapour pressure pwv of 14 mbar,which equals approximately a relative humidity of 45%. However,the same MF1 foil has a WVTRA of 0.360 g/(m2 day) at 25 8C and75% RH. The general influence of the relative humidity on theWVTRtot/A is expressed in Table 2.

Compared to MF1, an AF foil (see Fig. 7) has a typical watervapour transmission rate WVTR that is a factor 10 smaller: 0.001 g/(m2 day) at 25 8C and ambient pwv. The measured pressure willalways consist of both pdry air and pwv, but only at high relativehumidity will pwv have a certain influence on the total pressure(see also Table 3).

3.3.3. Dependency of WVTR and GTR on temperature

For both the WVTR and the GTR can the dependency ontemperature be written as an Arrhenius type of equation [2,52,53]:

QðTÞ ¼ QðT0ÞeððEa=RÞðð1=T0Þ�ð1=TÞÞÞ (43)

where R is the gas constant and Ea the activation energy. The valuesfor this activation energy are given in Table 4 for every type of VIPfoils. An analogue table as Table 2 can be written to show theinfluence of the temperature on the (WVTR and the) GTR [23]:

Later will be shown that, as consequence, the service life of avacuum insulation panel will depend on the same Arrhenius typeof equation if the water vapour saturation pressure curve isassumed to be linear in a certain temperature interval.

The values of all these properties are put in Table 4 for the threelayer materials AF, MF1 and MF2 (see Fig. 7) for which measure-ments are done in [23,25].

It becomes clear that MF1-covered VIPs have a high pressureand mass increase rate, which makes them unattractive for use as ahigh barrier envelope in vacuum insulation panels. The values of AFand MF2 foils are comparable and useful for VIPs. However, it must

ed that the values for the MF3-type of laminate have been retrieved at 23 8C and a RH

Barrier envelope materials

Foil AF Foil MF1 Foil MF2 Foil MF3

(30�2)�10�9 87�10�9 (14�2)�10�9 (3–8)�10�9

(3.2� 0.2)�10�9 8.1�10�9 (1.2�0.3)�10�9 /

(2.9� 0.2)�10�9 7.5�10�9 (1.1�0.3)�10�9 /

2.6� 0.2 7.2 1.0�0.1 /

0.23 0.59 0.09 /

26�2 40�7 28�3 /

0.0005� 0.0003 0.0067 0.0011�0.0004 0.001–0.002

Fig. 15. (a) Principle of the foil lift-off method with two evacuable spacers (Caps, 2008) and (b and c) measuring of the heat flux through an inner fleece with the thermal

method va-Q-check developed by va-Q-tec [73].


be said that AF foils have a much higher linear thermalconductivity as shown in Section 2, which makes the AF-foil lessattractive for vacuum insulation panel envelopes and the MF2-foilthe most promising of them all.

3.3.4. Measuring the internal pressure

Furthermore, the method of measuring the increase of gas andwater vapour pressure has been studied [26,54]. A distinction ismade between non-thermal methods and thermal methods.

A first technique was introduced based on the foil lift-off method:A vacuum insulation panel is completely or partially placed betweentwo evacuatable chambers (see Fig. 15[left]). The upper and lowerspaces are simultaneously evacuated and as soon the pressure in thechamber drops below the internal pressure of the VIP, the envelopeof the panel lifts off. However, this method is unpractical forpressures below 0.5 mbar [54]. A second non-thermal method isdeveloped for non-thermal testing [55,56] of VIPs applied in abuilding construction [26]. The method is based on RFID, the radiofrequency identification technique that receives data remotely fromso called RFID-tags incorporated in a product. Passive tags (tagswhich do not require an internal power source) and active tagstransmit information from an additional sensor which measures thetemperature or pressure. An analogue system is developed by ‘va-Q-tec’ to detect damaged VIPs: No signal of the tag will be received ifthe internal pressure has reached a certain value (e.g. 600 mbar). Thetags can only be used for VIPs with thin metallized foils, becausethick metal foils will completely shield the signal.

Furthermore, a thermal detector for gas pressure measure-ments in VIPs is developed by ‘va-Q-tec’ [54]. A thin fibre fleece isplaced below the VIP envelope above a thin metal plate. Thethermal conductivity of the fibre fleece is measured by placing a

transmitter/detector on it from the outside, introducing a heat flowthrough the envelope and the fleece to the metal plate. A steadystate heat flux is reached after a few seconds, measured by thedetector (see Fig. 15[right]). From this result, a value for thethermal conductivity of the fleece can be calculated, which isfunction of the internal pressure. The thickness of the fleece willdetermine the measurable pressure range.

3.4. Service life prediction for VIPs with a fumed silica core and foil

envelope

An analytical model for the service life can now be written withthe information of Section 3. Eq. (34) can be rewritten with

d pg

dt¼

p0QgðT;’; lÞVi

D p (44)

du

dt¼

QH2OðT;’;A; LÞmdry

D pH2O ¼u0

a’e

ða’e � uÞ (45)

where Dp is the difference in pressure between outside and insidethe panel and where Qg and QH2O can be defined with the WVTRand GTR of the VIP [52,2,51], As a result, we come to the proposedanalytical model of Tenpierik et al. [49]:

DlcðtÞ �@lc

@pgpg;eð1� e�t=tg Þ þ @lc

@ pwvpwv;eð1� e�t=tw Þ þ @lc

@u

� du

d’’eð1� e�t=tw Þ (46)

Fig. 16. Centre-of-panel thermal conductivity for VIPs with a fumed silica core

defined using Eqs. (46) and (47), for panel sizes of 50 by 50 by 1 cm3 and 100 by 100

by 2 cm3 and for three different foil types AF, MF1 and MF2 (see Fig. 7). The graph is

calculated with values for the envelope WVTR and GTR determined at 23 8C and 75%

relative humidity as mentioned in [51] (see Fig.13 and Fig.14), a porosity of 90%, a

dry core density of 200 kg/m3, u(w) as defined in Eq. (8), @lc/@u as defined in Eq.

(13), @lc/@pg as calculated from Eqs. (4) and (5), and psat as defined in Fig. 5, and

matches the short-term increase of thermal conductivity as defined by Schwab

et al. [51]. The influence on pwv is yet incorporated in the retrieved values from

measurements on the dependence on the water content u. The inner air pressure is

assumed to be zero at t0 = 0 years whereas commercial products have an inner air

pressure between 0.5 and 5 mbar and no getters and desiccants have been taken

into account. One assumes that the VIP envelope properties remain the same, i.e.

that no degradation will occur through time, which is unlikely (see Section 3.5).


where tg and tw are (temperature and relative humiditydependent) time constants, defined as

tg ¼eV

GTRðT;’ÞT0

p0Tand tw ¼

rdryV

WVTRðT;’Þ1

psatðTÞdu

d’(47)

where e is the porosity of the core material, V the total volume of thecore material, T0 a reference temperature, p0 a reference pressure,psat the water vapour saturation pressure of the air and where rdry isthe dry density of the core material. This increase of thermalconductivity as in Eq. (46) due to the intake of moisture and gas isexpressed graphically in Fig. 16 for three different foil types.

Eq. (46) can be adapted to the retardation due to the getters anddesiccants, that are added to the vacuum insulation core of fumedsilica, by replacing the term t with respectively (t � tget) and(t � tdes):

DlcðtÞ �@lc

@ pg

pg;eð1� e�ðt�tgetÞ=tgÞ þ @lc

@ pwv

pwv;eð1

� e�ðt�tdesÞ=tw Þ þ @lc

@u

du

d’’eð1� e�ðt�tdesÞ=tw Þ (48)

where tget and tdes are the time shifts due to getters and desiccants,calculated as

tget ¼cgetmget

ð pg;e � pg;i;0ÞGTRðT;’Þ and

tdes ¼cdesmdes

ð pwv;e � pwv;i;0ÞWVTRðT;’Þ (49)

Table 5Regression parameters for Eq. (51) [29]. The values obtained by Tenpierik et al. [49] with

Core Envelope a [m�(c+1) year] b [m K/W]

SiOx MF3 (5.17�0.80)�103 333�1

AF (3.61�0.39)�103 270

where cdes [kg/kg] is the maximum capacity of the desiccant, mdes

the mass of the added desiccants, cget [m3/kg] the maximumcapacity of the getter and mget the mass of the added getters.However, it must be noticed that Eq. (48) is only correct for t > tdes

and t > tget: If t is smaller than one of those (or both), thecorresponding term has to be equalled with zero. This means thatas long as the getters and/or desiccants did not yet reach theirmaximum capacity, the pressure and/or moisture content in thevacuum insulation panels will not yet start to raise.

With these equations given for the increase of the thermalconductivity of the core, an equation for the service life time tSL of avacuum insulation panel can be derived [49]:

DlcðtÞ �@lc

@ pg

pg;eð1� e�ðtSL�tgetÞ=tg Þ þ @lc

@ pwv

pwv;eð1

� e�ðtSL�tdesÞ=tw Þ þ @lc

@u

du

d’’eð1� e�ðtSL�tdesÞ=tw Þ (50)

If no getters and desiccants time shifts are assumed, the servicelife can be solved from Eq. (50). For VIPs with a core of fumedsilica, an analytical estimation can be made as in Eq. (51) with adifference of maximum 17% compared to the result of thedifferential Eq. (50) [49]:

tSL � aebðllim�l0�lwÞdviplvip

Svip

� �cT0

Te½ðEa=RÞðð1=TÞ�ð1=T0ÞÞ� (51)

where a [sm�(1+c)], b [m K/W] and c are regression parametersdefined in Table 5, dvip the VIP thickness, Ea the activation energy, R

the gas constant and lw the thermal conductivity of liquid waterand water vapour at equilibrium with ambient conditions whichcan be determined [2] as:

lw ¼ 0:029uð’eÞ þlwv;0

1þ ðP1=2=’e psatÞ(52)

The regression parameters a and b in Eq. (51) are parametersdepending on the combination of envelope and core material,while the parameter c is a quadratic regression function dependingon the temperature:

c ¼ c0 þ c1T0

T

� �þ c2

T0

T

� �2

(53)

where c0, c1 and c2 have to be derived from experimental results.All these parameters a, b and cx have been investigated [51,52] forpanel thicknesses between 10 and 50 mm, temperatures between�5 and 45 8C and for panels with a lvip/Svip in the range of 2–10 m�1

for MF3- and AF-covered VIPs with a core of fumed silica. Withinthis range, the parameters can be assumed to have the value asmentioned in Table 5.

With these values, a service life in the range of 18–40 years isfound for MF3-VIPs panels with a lvip/Svip between 2 and 10 m�1 at23 8C. For the same conditions, a service life between 60 and 160years is found for AF-VIPs. Analogue values are found according toolder studies and models [51]. In these models, similar values arefound for MF2-VIPs as the MF3-VIPs, while for MF1-VIPs anunacceptable service life of less than 10 years was found.

a regression analysis on the data of Schwab (2005c) and Simmler and Brunner [52].

Ea [J/mol] c0 c1 c2

50�2 15.4�2.6 �31.3�5.3 15.1�2.6

26�2 8.8�1.4 �17.8�2.8 8.2�1.4

Fig. 17. Sound reduction R (dB) of VISs with 15 kg/m2 facings and absorption

coefficient a of a VIP under vacuum condition as function of frequency (redrawn

from [60,61]).


3.5. Other ageing factors

Other climate factors can age building materials and buildingcomponents. These climate factors may be divided into ultraviolet,visible and near infrared solar radiation, ambient infrared heatradiation, temperature changes or freezing/thawing cycles, water(e.g. moisture, relative air humidity, rain and wind-driven rain),wind, erosion, pollutions, micro-organisms, oxygen and timedetermining the effect for all the previous mentioned factors towork [57]. The ageing factors on polymer degradation for VIPenvelopes are expressed by Thorsell [58], divided in four maingroups: The chemical environment, thermal shocks, ultravioletlight and high energy radiation. However, the real effect ofchemicals and radiation on VIPs is unknown and it is unclear whatis meant with ‘high energy radiation’ by Thorsell [58] in thiscontext. One possibility might be ‘shortwave visible radiation’which is the visible radiation with the highest photon energy, alsotaking part in the degradation of several materials (e.g. [59]).

Furthermore, accelerated ageing of the envelope material byhigh temperatures and damaging by mechanical loads is men-tioned. These mechanical loads can be a nail penetrating theenvelope with immediate pressurization as consequence or aworker stepping on the VIPs on the construction site without anyvisible damage. A VIP is such a fragile material that one has toassume proper and careful handling at both the manufacturing andconstruction site. The best way to ensure safe handling would be tonever allow the bare VIPs at the construction site but supplyingthem integrated into more durable building components [100,58].

4. Acoustical properties of applied VIPs

The hygro-thermal properties of vacuum insulation panels havebeen previously discussed in Sections 1–3. A large potential indecreasing the thickness of the thermal layer due to the highinsulation value has been noticed here. However, also the soundinsulation properties of VIPs are important if they are used asbuilding insulation. Research on these properties started by Lenzet al. [60] and was continued by Cauberg and Tenpierik [61] andMaysenholder [62]. However, there is not much known on theacoustical properties besides these three articles.

The acoustical knowledge or properties of VIPs can be divided inthree groups: The acoustical properties of a single vacuuminsulation panel, the properties of with VIPs insulated sandwichelements and the properties of insulated massive walls with VIPs.Important here for are three factors which determine theacoustical quality. These are noise banning and sound absorptionfor facades and the dynamic modulus for sprung floors.

4.1. Acoustical properties of a single VIP

The sound insulation of a material or structure is in the firstplace determined by the law of mass, in which the weighted soundreduction index Rw is expressed as Rw � 20 log(1 + 3.4m) where m

[kg/m2] is the mass per area. With a bulk density of 2.2 kg/m2 for10 mm VIPs and 5.7 kg/m2 for 33 mm VIPs, a theoretical weightedsound reduction index in the range of 19–26 dB can be found [62].An other important property is the E-modulus: A value of 0.25 MPafor a VIP is found according to the international standard ISO 9052-1:1989, while a value of 0.49 MPa is found for the core materialfumed silica.

The same acoustical performance is found in measurements byMaysenholder [62] while much lower values are found forperforated VIPs. The general behaviour of single VIPs is heredescribed as ‘‘according to mass law, followed by a coincidencedip’’. The weighted sound reduction index Rw drops with 10–15 dBat this coincidence frequency. The aim in building acoustics is not

to have this coincidence dip in the range of traffic noise (�1 kHz),but it is found for VIPs with a fumed silica core in a disturbing zonebetween 1 and 2 kHz [60,62] (see Fig. 17).

It becomes clear that due to the strong reduction in thickness(and as consequence also due to the high thermal properties) theacoustic performance will have to be corrected with thicker andheavier facings in vacuum insulated sandwich elements.

4.2. Acoustical properties of VIP insulated structures

4.2.1. Vacuum insulated sandwiches VISs

The acoustical properties can be improved by applying VIPs insandwich elements, a common application of vacuum insulationpanels. As result, the Rw value will increase due to a higher massper surface, but dilatational resonances will occur due to mass-spring-mass system with the VIP as the spring and the facings asmass.

These VISs have been investigated: Sandwich elements withTRESPA facings by Cauberg and Tenpierik [61] and with steelfacings in Maysenholder [62]. Comparable data are found in bothstudies: A weighted sound reduction index Rw � 28 � 1 dB and avalue of 4 dB lower for traffic noise. This value Rw can be improvedstrongly by adding two rubber layers of 1 mm between the (metal)facings and the vacuum insulation panel: The coincidence frequencywill move out of the disturbing zone to 3 kHz and the weighted soundreduction index will follow the mass law Rw = 34 dB.

4.2.2. VIP insulated cavity wall

VIPs applied as cavity insulation in a massive wall will notimprove the acoustical performance of the wall: The vacuuminsulation panels have a sound absorption coefficient of 0.0 up to500 Hz and an absorption coefficient of 0.2 � 0.05 at higherfrequencies, as determined by Cauberg and Tenpierik [61].

5. Building applications of VIPs

One main benefit with VIPs is the reduction of the requiredthickness of the insulation layers compared to traditional thermalinsulation materials in building applications due to the much

Fig. 18. Comparison [left] between the thermal conductivities of conventional and advanced insulation materials and solutions (http://gfp.lbl.gov/) and [right] between the

required thicknesses for conventional insulation (e.g. glass wool) and a VIP [72].


lower thermal conductivity of VIPs, i.e. approximately 5–10 timeslower. As a result, the necessary building insulation thickness maybe decreased with a factor of 5 up to 10, which is depicted in Fig. 1and Fig. 18. However, for the application of VIPs, one has to payattention to certain disadvantages compared to the conventionalinsulation materials (see Table 6).

Vacuum insulation panels are already introduced to the marketin large-scale production. However the manufacturing is stillmainly hand-labour. The last decades, several building applica-tions for VIPs have been proposed and/or tested and a large-scalestudy has been carried out on the possibilities of VIP insulatedbuilding envelopes [63]. Example building applications in papersmay be divided in several groups:

A first application method consists of insulating existingbuildings on the inside [33,34,97] or on the outside [64–68,100]of the existing building envelope. Installing additional insulationon the inside results mostly in a great loss of floor space, thereforeVIPs are of interest for renovations. Special attention has to be paidon low surface temperatures and possible condensation damagesat connections to surrounding compounds.

A second current common application method is applying VIPsas vacuum insulated sandwich elements (VISs) in door andwindow frames, in curtain walls and in non-load bearing walls[42,64,69,65,70,45,66,63,71,68]. In the studies on window anddoor frames, an improvement up to 50% of energy transmittance isnoticed for the overall frame system if the performance of theoverall system is not comprised by an adapted frame design.

A third application group can be found in insulating flat roofs,loggias, terraces and internal floors, where the small thickness ofthe needed insulation layer enables a simple construction for astepless transition between the interior and exterior space[64,69,72,65,73,53,68] (see Fig. 19). Measurements have been

Table 6Summary of the advantages and disadvantages of vacuum insulation panels for buildin

Advantages

High performance thermal insulating.

Increased floor area.

Appropriate refurbishing of existing buildings

with high restrictions.

Can be used as ‘thermal-bridge-insulation’

on places where not much space is available.

Up to now the only material for which 4 cm is

enough to meet (most of) national building regulations.

Lower operating temperature increases thermal

performances, which is beneficial in Nordic countries.

performed on VIP insulated flat roofs and have shown that a servicelife of several decades is very likely. It has also been shown that themoisture load of flat roofs can strongly influence the effectivethermal conductivity.

A fourth group of building applications is VIPs as main insulationof the building envelope for new buildings [65,63,22,52,53].Adapted wall constructions are shown/proposed for integratingVIPs in a sustainable solution [72,74] to come to average thermalperformances of Uwall � 0.1 W/(m2 K) with limited thicknesses. Insuch applications, it is always studied whether or not the completeenvelope building components should be pre-assembled inadvance, to ensure the proper handling of the VIPs.

A last group of VIP building applications can be found inbuilding installations, e.g. pipe insulation [75], insulation for hotwater cylinders [76] or under-insulation of floor heating [77,78].

6. Other possible high performance thermal insulatingmaterials and solutions

Vacuum technology is only one way to achieve a highperformance thermal insulating solution. As comparison, a shortextension towards gas-filled panels (GFPs) and aerogels will bemade. Both GFPs and aerogels are new high performance thermalinsulating solutions and materials developed in the last decade,introducing new technologies or possibilities which may be usefulin the development of such new high performance thermalinsulation materials.

6.1. Gas-filled panels

Gas-filled panels (GFPs) try to minimize all parameters of heattransfer by using a low-conductive gas as the main insulator to

g purposes.

Disadvantages

Very fragile and protection for puncture of the foil is necessary.

Reducing thermal performances through time.

Limited service life may require replacement.

Effective thermal performance will be reduced by thermal bridging

of the VIP envelope.

Increased structural thermal bridging due to the envelope.

Less suitable for traditional timber wall structures.

Inflexible: On-site adaptations (cutting of panels) are not possible.

Production inaccuracy of the panel sizes

http://gfp.lbl.gov/

Fig. 19. [Left] Comparison between a vertical cross section of a connection of a conventional insulated terrace and a construction with a comparable with VIS insulated terrace

construction [65] and [right] an example of a slim balcony construction [73].

Fig. 20. View of the barrier foil and the baffle structure inside a gas-filled insulation panel (http://gfp.lbl.gov/).

Table 7The effective thermal conductivity for an optimum number of cavities, based on the

lowest marginal cost of the fabrication of the gas-filled panels [109].

Panel thickness Effective thermal conductivity, le [W/(m K)]

Air Argon Krypton

25 mm 0.0350 0.0213 0.0106

50 mm 0.0380 0.0226 0.0106


influence both the gaseous thermal conductivity lg and thethermal conductivity of the solid structure lcd. Because the gas isused at ambient pressure, a core material is no longer necessary totransfer forces as in vacuum insulation panels. To enclose the gasand to decrease the heat transfer due to radiation, GFPs use a low-emissivity surface baffle structure in a barrier envelope.

The thermal conductivity through the gas is the most importantheat transfer in a GFP. High performance GFPs use gases whichhave a lower thermal conductivity then air. Two general rules [79]can be expressed to explain this lower thermal conductivity of agas. Firstly, a gas with a higher molecular weight will have a lowerthermal conductivity. Secondly, a mono-atomic gas (e.g. argon,krypton and xenon) will have a lower thermal conductivitycompared to polyatomic gases with the same molecular weight.The reason for this is an extra possible energy transfer due torotation and vibration of the gas molecules. Most attention is paidto the potential of GFPs with the inert noble gasses argon, kryptonand xenon as a gas-fill because these gasses can be extracted fromthe atmosphere and as consequence have a global warmingpotential of zero.

Gas-filled panels are made of two types of polymer films:Metallized films are used in a tied assembly called ‘the baffle’,producing a cellular structure in the panel which suppressconvection and radiation (see Fig. 20). Low diffusion gas barrierfoils are used in a hermetic ‘barrier’, keeping the panels gas-tight.Depending on the type of foils used for the structure, these panelsmay be made as stiff or flexible insulation panels. As a result of thestructure, the total panel will have a thermal conductivity close tothe still-gas thermal conductivity of the fill (see Table 7).

The barrier is a hermetically sealed enclosure to maintain thegas-fill and a critical component for the GFPs. The foil will have toact as an effective gas barrier in two directions: Moisture and airgases are driven into the panel and the gas-fill is driven out of the

panel. Similar to the envelope foil for vacuum insulation panelswill the quality of GFP barriers be quantified by their gastransmission rate. However, a distinction will be made for eachtype of gas because of the specific gas content of the panel. A gasloss of 0.1 vol% is acceptable for GFPs. Within this limitation, avariety of films and foil configurations are commercially available.A first group includes all foils of vacuum applications with verythin layers of aluminium such as vacuum insulation panels. Asecond group is polymer barrier resins, including ethylene-vinyl-alcohol, polyvinylidene-chloride and polyvinyl-alcohol [79].

The baffles are necessary to suppress convection and radiation.Making compartments by constructing the baffle out of a solidmaterial minimizes the convective heat transport of the gas. Forthis gas division into compartments, thin sheets are assembled in athree-dimensional form of multiple layers of cavities. The radiationheat transfer is decreased by using low-emissivity cavity surfaces,which are inexpensive and available as metallized thin polymerfilms. The dimensions of the foil and the formed cavities areselected to minimize the gas convection in these cavities, to reducethe solid conduction and to minimize the economic cost. There isan optimal number of baffle layers [80] to use in a GFP with itsspecific thickness, gas-fill and temperature difference regardingthe economic cost and the thermal performance.

http://gfp.lbl.gov/

Fig. 21. Structure of the nano-porous SiO2 network of silica aerogel [14] and a Scanning Electron Microscopy SEM image of the structure of clay aerogel [13].


6.2. Aerogels

The remarkable properties of aerogel result from its extra-ordinary physical and chemical structure. Silica aerogels consist ofa cross-linked internal structure of SiO2-chains with a largenumber of air filled pores. These pores of aerogel are very small:Pure aerogel has an average pore diameter between 10 and 100 nm[81]. However, aerogels in general will have pore sizes between 5and 100 nm, depending on the purity and the fabrication method[82]. These pores will take 85 up to 99.8% of the total aerogelvolume. Due to its extraordinary pore sizes and the high porosityachieves aerogel remarkable physical, thermal, optic and acousticproperties. However, this also results in a very low mechanicalstrength. The high porosity makes aerogels the lightest solidmaterial know at present: However it has a skeleton density of�2200 kg/m3, the high porosity results in a bulk density up to aslow as 3 kg/m3 whereas, e.g. the density of air is 1 kg/m3 [83–85].

Silica aerogels have an overall thermal conductivity of�0.017 W/(m K) at ambient pressure. The solid thermal conduc-tivity of dense silica is relatively high, but silica aerogels have sucha small fraction of solid silica. Furthermore, the inner skeletonstructure has many ‘dead-ends’ resulting in an ineffective andtortuous path of thermal transport. Lower density aerogels can bemade to decrease the solid conductivity, but this may result inmechanically weaker properties and in wider pores, i.e. lessoptimal thermal properties (see Fig. 21).

The low gaseous thermal conductivity may be explained usingEq. (6): Silica aerogels have both an intrinsic low characteristicpore size d and a very high porosity. As a result, the gaseousthermal conductivity will have a large influence on the overallthermal conductivity of aerogels, but will at the same time bestrongly reduced due to the Knudsen effect at ambient pressure. Asbrought up before, the gaseous thermal conductivity can be furtherreduced by filling the aerogel with a gas, by decreasing themaximum pore size and by applying a vacuum in the aerogel. Anoverall thermal conductivity of 0.008 W/(m K) can be reached forsilica-aerogels by applying a pressure of �50 mbar or less if noattempts are made to decrease the radiation transfer [81,84–87].

Silica-aerogels are reasonable transparent in the infrared. Theradiative transfer will become a dominant factor of thermalconductivity at higher temperatures (>200 8C), but will representno problem at low temperatures. Radiative transfer can besuppressed by adding an additional component to the aerogel(before or after the critical drying) that either absorbs or scattersinfrared radiation, e.g. carbon black. In this way, the overallthermal conductivity at ambient pressure can be decreased to avalue of 0.0135 W/(m K) at ambient pressure and to 0.004 W/(m K)at a pressure of �50 mbar or less.

Silica aerogels have interesting optical properties. It has a hightransmittance of radiation within the range of visible light.However, reflected light appears bluish and transmitted light

appears slightly yellow. The scattering of the light can be explainedby bulk or Rayleigh scattering and by exterior surface scattering.This Rayleigh scattering is caused by interaction with in-homogeneities in solids, liquids or gases (e.g. dust particles inatmosphere) and becomes more effective when the size of theparticles is similar to the wavelength of the incident light (380 upto 780 nm for visible light). The presence of a certain number ofpores within this range can act as so-called ‘scattering centres’. Theefficiency of scattering will depend on the size of the scatteringcentres, while different wavelengths of radiation will scatter withdifferent magnitudes. A heating-treatment of the aerogels canincrease their transparency furthermore with�6% to a value of 76%[102] due to water desorption and burning of organic components.Furthermore, the optical properties can be influenced by para-meters of the sol–gel process by selecting optimal synthesisparameters [83,87].

7. Beyond vacuum insulation panels

The reduction of the gaseous conductivity of a material usingvacuum technology and low conductivity gases has meant a bigleap in thermal insulation. While the traditional insulationmaterials developed before the 1960s had a theoretical minimumthermal conductivity equalling the thermal conductivity of air, thenew developed insulation materials have initial thermal con-ductivities up to 10 times lower than this limit.

But also for these new high performance insulating solutions,strong restrictions were found: Although current vacuum insula-tion panels seem highly promising with their thermal conductiv-ities between 0.004 and 0.008 W/(m K), it became clear that VIPsare a very delicate system. Damaging the envelope can severelyand rapidly increase the thermal transmittance of a large area.Such a process will slowly occur anyway because of gas and watervapour intake through time, while also the need for metals in thecommon VIP envelope increases the overall thermal conductivity,compared to the remarkable low centre-of-panel conductivity ofVIPs.

7.1. Possible improvements for current VIPs

Based on previous sections, some possible improvements couldbe brought up to reduce the effective thermal conductivity or toprolong service life of currently common VIPs. The development ofa core material with a lower solid core conduction could bring thecentre-of-panel thermal conductivity of a VIP further down. Alsothe development of a polymer envelope material that no longerneeds metal layers to maintain the inner lower pressure may resultin lower effective heat transmittances. The development of ahigher quality of air and moisture tight envelope could extend theservice life of current VIPs for the same defined criteria, while also acore material with smaller pores could prolong the service life,

Fig. 22. Closed-porous materials may make it possible to avoid the envelope


when a high quality vacuum is no longer necessary due to theKnudsen effect.

Such improvements are likely to occur in the coming decade(s),but it might be questionable if they would mean a new big leapforward in high performance thermal insulating materials.

7.2. New high performance thermal insulation materials

To come to such a new big leap in high performance thermalinsulating materials, different question may be asked:

necessary to maintain the inner vacuum in the material, i.e. making a so-called

vacuum insulation material (VIM).
- Is it possible to develop a vacuum insulation material or gas
insulation material based on the same principles but without its

disadvantages?

- M
ay we develop a material that does not need an envelope or that is
not affected by gas and water vapour intake over time?

- M
ay we combine both the technology of VIPs and GFPs to achieve a
more robust high performance thermal insulating material?

Based on the above asked question, a possible division may bemade in potential new high performance insulation materials: Afirst group can be found in core-based materials with no envelope

using vacuum-technology to achieve high performance thermalproperties, resulting in a material that can be cut, placed andhandled on-site like traditional insulation materials, e.g. mineralwool, expanded polystyrene (EPS) and extruded polystyrene (XPS).A second group can be found in envelope-based solutions with no

core, i.e. a surrounded vacuum space. Such solutions, e.g. as aspecial case small vacuum spheres, may then be applied as loose-fill insulation in building envelopes. Both groups of potential VIMsdeal on their own way with the current limitations in vacuumtechnology in order to achieve a high performance thermalmaterial or solutions. Both options will be questioned in the nextsections.

One of the main issues for common VIPs is the vacuum-maintaining envelope around the core material, putting stronglimitations on the application of it in building envelopes. Wheretraditional insulation materials can be ordered en-masse, theenvelope of VIPs makes it necessary to pre-calculate and fabricateevery single vacuum panel since on-site adaptation is impossible.Can we develop a high performance thermal insulation materialthat no longer needs this envelope, leaving only the core materialwhile still applying vacuum technology to come to a highperformance thermal insulation material? Three possibilities tocome to such a material can be concluded [96] from previousSections 1–4: A first possibility can be found in the poreconstruction. The core material of current VIPs is 100% openporous, where a possible solution can be found in porous materialswith 100% closed pores, maintaining the vacuum on the level of thepores. A second possibility may be found in the so-called ‘Knudseneffect’, where the need of a vacuum is expressed as function of thepore sizes. Influencing the pores sizes and pore size distributioncan lead to the need for a vacuum of much lower quality or evenevolve in a material that no longer needs a vacuum but that is stillbased on the principles of vacuum technology. A third possibilitymay be found in the combination of both possibilities. All threepossibilities will be addressed.

Another option was proposed combining the structure of GFPswith vacuum technology, leaving the core material behind andresulting in only a stiffened vacuum envelope. In this way, the twomain factors (solid thermal conduction lcd and gaseous thermalconductivity lg) are almost completely excluded, leaving onlyradiation as the main heat transfer. However, the structure of GFPsresult from the fact that gasses are applied at atmospheric pressureand that no forces have to be transferred, where in VIPs the forcesup to 1 bar (100 kPa) due to the vacuum are transferred by the core

material. A solution can be found in the shape of the vacuumenvelope: It is commonly known that spheres are most resistant tovacuum or high gas pressures in such a way that an internalpressure of 1.5 mbar is no problem. However, these solutionsconsist mostly of heavy metal spheres to prevent the sphere fromcollapsing. They can still form a solution if the size of the spherescan be kept in the range of millimetres or micrometres to reduceforces in such way that a low-conductivity material can be used toform the spheres. Such spheres can then be applied in buildings asloose-fill insulation or repacked in panels.

7.3. Vacuum insulation materials (VIMs)

As previously brought up, a closed-porous material could have ahigh potential as possible vacuum insulation material or VIMwhich no longer requires a panel envelope to maintain an innervacuum. Such a vacuum insulation material (VIM) can be definedas a material that is basically homogenous and which achieves its high

performance thermal insulating qualities due to vacuum technology in

its closed-porous structure (see Fig. 22).A first requirement for such a VIM are pores closed for air which

maintain a vacuum on the scale of the pores and making anenvelope no longer necessary. This would result in a material thatcan be cut on-site and for which damaging by perforation with anail only would lead to a local thermal bridge. On such places, onlya certain limited amount of pores will be ventilated, instead of awhole large region as for VIPs. A closed-porous material may evenhave other advantages compared to open-porous materials: Anopen-porous structure could result in capillary condensation ofwater after a long exposure to humid circumstances. Here,materials with closed pores are more desirable because of itsuseful properties as mechanical strength and resistance tohumidity [88].

However, one of the restrictions stated for VIP core materials isa 100% open-porous structure. This is necessary to be able toevacuate the material. A closed porous material where intake of airdoes not occur through time, means on the same time that avacuum cannot be achieved in the pores by evacuation as incommon VIPs because both are based on the same principles.Solutions for this problem may be found in the production of thematerial under vacuum conditions (or only for the productionstage where the closed-pores are formed) or in chemical binding ofthe air in the pores.

The synthesis of closed porous silica has been studied recently[88,89] in order to develop an increased resistance to humidity andmechanical strength. The closed-pores were synthesized byincorporating rapid drying with the sol–gel process and the effectof the drying temperature on the closed porosity of the silica hasbeen studied.

‘‘Closed porous silica were synthesized (. . .) from starting agents

with the molar ratio of 0.04 TEOS, 1 H2O, 0.022 ethanol and 0.001

HCl. These were mixed and stirred at room temperature until a

Fig. 23. Dependence of closed porosity on the drying temperature for silica

(redrawn from [89]).

Fig. 24. Manufacturing materials with a low-conductivity solid state lattice filled

with tiny nano-pores may make it possible to avoid the envelope necessary to

maintain the inner vacuum in a VIP (as a vacuum is no longer necessary to reduce

the gaseous thermal conductivity to a minimum), resulting in a nano insulation

material (NIM) with an open pore structure. In addition, nano-structures may also

be used to make a closed-porous NIM with or without vacuum in the pores.

Fig. 25. Pore size distribution of MCM-41 (redrawn from [90]) and of (mesoporous)

calcinated products (redrawn from Kuroda [110]) with even smaller pore sizes. The

development of mesoporous materials do not seem to be a problem with current

technology, though no information is found on the thermal conductivities of these

materials.


transparent solution was obtained after the completion of

hydrolysis of TEOS in the acidic condition. Then pluronic1 F127

is dissolved (. . .) with a weight fraction to that of water of 0.1. The

solution was rapidly dried (. . .) and a complete desiccation was

carried out using a drying over for a few hours. The final

calcinations was carried out at 600 8C for 5 h in air to completely

remove the copolymer as the structure-directing agent.’’ ([88], pp.663–664)

The effect of closed pores can be forced by controlling thetemperature of the rapid drying. Higher temperatures, which causemore rapid drying, are expected to promote the formation of theclosed pores due to the more instant solidification (see Fig. 23). Sofar, silica with a closed-porosity up to 48% has been obtained at adrying temperature of 200 8C [89]. However, it can be stated thathigher temperatures above 150 8C also promote the formation ofthe open pores [88]. However the development of partial closed-porous silica do not seam to be a problem with current technology,no information is found on full-closed-porous materials or theapplication of vacuum within these pores.

7.4. Nano insulation materials (NIMs)

A nano insulation material or NIM may be defined as a material

that is basically homogeneous and which achieves high performance

thermal insulating qualities mainly due to its open or closed nano-

porous structure.The principle of vacuum insulation panels is based on the

Knudsen effect, generally describing that the gaseous thermalconductivity lg can be influenced by controlling both the meanfree path lmean of the gas and maximum pore size d of the medium.While reducing the gas pressure in a material, the gas conductivityremains almost unaffected until the mean free path of the gasmolecules reaches values in the same order of size as the largestpores in the medium. When the pore diameter of the materialbecomes less than the average free length of path of gas molecules,the air molecules will only collide with the pore surfaces withouttransferring energy by this elastic impact.

It could be stated that a certain vacuum is no longer required if amaterial is used with a maximum pore size much smaller thanlmean of free air. Such so-called ‘mesoporous materials’ with a poresize d of 50 nm or less or ‘microporous materials’ with a pore size dof 2 nm or less, could reduce the gaseous thermal conductivitywithout applying a vacuum, thus achieving a nano insulation

material (NIM). A gaseous thermal conductivity of 0.2 mW/(m K)can be found for pore sizes of 2 nm according to Eq. (4).Furthermore, degradation of the overall thermal conductivitybecause of gas penetration can be neglected because air is allowedin the pores and an air-tight envelope is no longer necessary(see Fig. 24).

Aerogel is one of the materials that has a low thermalconductivity at ambient pressure partly because of the Knudseneffect. Common aerogels have mean pore sizes between 10 and40 nm, but will furthermore also have pores with sizes up to100 nm. The development of such a new material can be positionedat the branch of nanotechnology, a technology generally describedas ‘dealing with structures of 100 nm or smaller’.

Most well-known mesoporous materials are aluminium oxideor oxides of the metals 41Nb, 73Ta, 22Ti, 40Zr and 50Sn with highthermal conductivities. However, a mesoporous silica was devel-oped in the early 1990s by the oil company ‘Mobil’, resulting inmaterials called MCM-41 with pore sizes between 1.5 and 10 nm[90] and SBA-15 with pore sizes between 4.6 and 30 nm [91] (seeFig. 25). Since then, much research has been done on thedevelopment of mesoporous materials, proved by the appearanceof the journal ‘Microporous and Mesoporous Materials’ of Elsevier.The pore size of the materials can be modulated during thesynthesis by controlling the reaction time and temperature, by


using swelling organic molecules, by adjusting the surfactant, bypost-synthesis or changing the calcinations conditions [92].

As shown can both the Knudsen effect and a closed-porousmaterial have a potential as high performance insulating material.However, a 100% closed porous material is currently unlikely andas a result, intake of air and water vapour will occur as in thecurrent VIPs due to the pressure differences. The combination ofboth ideas behind VIM and NIM could result in a material that willdegrade through time, but prolonged in service life because thedriving forces are reduced by the need for a low-quality vacuumresulting from the Knudsen effect. It was brought up before inSection 7.3 that (partially) closed-porous silica has been developedby Pei et al. [88], Pei et al. [89]. It was not mentioned that thesedeveloped materials where mesoporous and that (partially)closed-porous silica with a pore size 7 nm have been developed.However, a contradiction may be concluded: Rapid drying willincrease the amount of closed pores, while in addition, rapid dryingwill result in the formation of larger pores. Although, the formedpores were still of moderate small size (12 nm).

7.5. Gas insulation materials (GIMs)

Similar as for VIMs can a gas insulation material (GIM) bedefined as a material that is basically homogenous and which

achieves its high performance thermal insulating qualities due to a low

gas-fill in its closed-porous structure.The idea was brought up previously to apply a vacuum by

binding the molecules of the present air. However, it wasmentioned that air consists mainly of triple bond nitrogen, whichis almost inert. Replacing the air molecules with other gasmolecules during the production process of the material canmake it possible to chemically bind these molecules (with the solidmaterial) with vacuum as a result. However, the same problemmay appear here: Most low-conductive gasses are noble gassessuch as argon (see Fig. 26), krypton or xenon, which may make itimpossible to bind them. Other (non-inert) gases are needed here.If no vacuum is aimed at and the low gaseous conductivity is aimedbased on the Knudsen-effect, replacing air molecules with a lowthermal conductive gas can result faster in a high performancethermal insulation material.

7.6. Dynamic insulation materials (DIMs)

Previous discussed possible solutions may lead to an insulationmaterial with a very low thermal conductivity as based on thestatic definition of thermal insulation. But can the future of high

performance thermal insulation materials be found in such a static or

single-value definition? Is it true that high thermal resistance is of

great importance for building envelopes? Introducing insulation

Fig. 26. Gaseous thermal conductivity of krypton (mW/(m K)) as function of the

materials characteristic pore size and the gaseous pressure at a temperature of

300 K. The values have been retrieved from Eqs. (4) and (5) representing the

Knudsen effect. Compare with Fig. 3 (air).

layers with a high thermal resistance has strongly reduced theenergy consumption during the heating season. However, anincreasing use of home electrical equipment in combination with ahigh thermal resistance and a higher air-tightness of the buildingenvelope have introduced the problem of cooling in manybuildings. The high thermal resistance of the wall may in thecooling season indirectly be a reason for an increased energyconsumption.

Building envelopes with variable properties seem in this pointof view much more favourable to decrease the overall energyconsumption of a building. Adapting the thermal performances tothe (changing) environmental conditions and user needs can leadto energy savings. A similar approach is already noticed in buildingglazing: Where first only the U-value of a window was of greatimportance, this has changed by a desire for dynamically regulateterms as the solar factor, the visible transmittance, etc. leading tothe development of electrochromic windows. Such windows areable to strongly control the daylight illuminance and the solar heatgains, resulting in stronger control of the cooling loads, the use ofelectrical light, comfort conditions, etc. [93,59]. It is maybe evennot their annual energy impact that is of most importance, buttheir performance at peak times resulting in more economicalperforming and downsized heating, ventilating and air-condition-ing (HVAC) systems.

Such a dynamic insulation material or DIM may be defined as a

material which has controllable thermal insulating properties within a

desirable range.

7.6.1. Dynamic vacuum

One question now is whether dynamic properties may also beintroduced for insulation materials. Or in other words, is it possible to

come to an insulation material with a controllable thermal conductiv-

ity? And which advantages has such a material compared to static

values? Controllable thermal conductivity would suggest that wecan control a certain (or multiple) mechanism(s) of heat transfer insuch way that we can vary the total heat transfer through thematerial within a wide range. In theory, this varying of thermalconductivity is something that is already achieved and discussed inthis review: By controlling the internal pressure, we are able to varythe gaseous component of the overall thermal conductivity of thematerial. In this way values between 0.004 and 0.020 W/(m K) maybe obtained for fumed silica, or even higher values if inneroverpressure is allowed. However, controlling the inner pressurein common VIPs would be unwieldy: Ventilating and evacuatingsuch a large volume requests much electrical power. This stands instrong contrast with the principles of electrochromic glazing, wherethe control of irradiance is achieved with a minimum of electricalpotential. Is such an economic process also achievable for varying the

thermal conductivity based on vacuum technology?

A solution may be once more found in a so-called closed-porousvacuum insulation material VIM with chemical binding of thegasses. What if the chemical binding of the inner gasses (with the solid

structure) can be achieved by applying a relative small electrical

potential? What if the balance between binding and releasing the gas

molecules can be achieved within a small range of different electrical

potentials? However, it can be questioned now whether chemicalbinding of the gasses is necessary. What if we, instead of chemically

binding the gas molecules, make them immobile for heat transfer?

Electrically charged gas molecules could be immobilized byapplying a potential, making them physical attracted by theopposite charged pore structure without chemical binding.

Such a dynamic vacuum has been yet achieved in the nineties(Patent DE 196 47 567 A 1) [103] by controlling the reversibleproperties of the applied getters and desiccants by applying anelectrical potential, resulting in an electric consumption of about5 W/m2.

Fig. 27. Step-like variation of the reflectance R in an ideally electrochromic material

as function of the wavelength l (nm) (redrawn from [94]). The plasma wavelength

may be defined as lp = (2pc/qe)(mee0/ne)1/2, where c, qe, me, ne and e0 denote the

velocity of light, the electron charge, the free electron effective mass, the free

electron density and the dielectric coefficient of vacuum, respectively.


7.6.2. Dynamic emissivity of inner pore surfaces

It was found that the overall thermal conductivity could becontrolled by adapting the inner gas pressure in the material.However, the gaseous thermal conductivity is only one term of theoverall performance. Another possibility may be found incontrolling the emissivity of the pore surfaces. However, in orderto obtain reflection modulation of a material, the material has to becrystalline so injected or extracted electrons can be free. Changingthe free electron density changes the plasma wavelength lp,making it possible to regulate the reflectance. Increasing theelectron density results in an increasing lp, moving the reflectanceedge to lower wavelengths (see Fig. 27) and vice versa. Materialswith a high free electron density will be high reflecting materials,while a material with a low free electron density has a lowreflectivity [94].

7.6.3. Dynamic solid core thermal conductivity

One might also envision to be able to regulate the solid corethermal conductivity which may contribute to a fully dynamiccontrollable vacuum insulation material. Possible ways to accom-plish this may be by application of electrical fields in order tochange the structure, electron density, chemical bonds, etc. in thematerial lattice. The fundamental understanding of the thermalconductance has to be explored in order to achieve such a dynamicsolid core conductivity.

7.7. New thoughts and ideas

The previously mentioned list of ideas ‘beyond VIPs’ is meant asa feasibility study for future insulation materials and is notexhaustive. The brought up concepts are still based on a classicpoint of view where thermal insulation materials mainly only haveto fulfil their function of reducing heat transfer through a buildingenvelope, while many other ideas, e.g. load bearing insulationmaterials, could be brought up to come back to ‘‘the buildingenvelope as the climate regulating skin it once was’’ [95].

8. Conclusions

Vacuum insulation panels (VIPs) in their current form can resultin a strong reduction of energy consumption if they are properlyinstalled. It was even proven [98] that the requirements of theKyoto-protocol are easily met if every non-insulated building

currently existing in the European Union is insulated with a layerof only 2 cm thick VIPs. However, the proper application of VIPs isproblematic because of factors as fragility, no possibilities foradaptation at the building site, high cost, thermal bridges anddecreasing thermal properties through time. It might also bedifficult or challenging to introduce new solutions within existingbuilding traditions.

Concepts for advancing beyond VIPs are investigated, wherepossible solutions such as vacuum insulation materials (VIMs) andnano insulation materials (NIMs) are introduced.

Acknowledgements

This work has been supported by the Research Council ofNorway, AF Gruppen, Glava, Hunton Fiber as, Icopal, Isola, Jackon,maxit, Moelven ByggModul, Rambøll, Skanska, Statsbygg andTakprodusentenes forskningsgruppe through the SINTEF/NTNUresearch project ‘‘Robust Envelope Construction Details for Buildings

of the 21st Century’’ (ROBUST).

References

[1] X. Wang, N. Walliman, R. Ogden, C. Kendrick, VIP and their applications inbuildings: a review, Proceedings of the Institution of Civil Engineers. Construc-tion Materials 160 (CM4) (2007) 145–153.

[2] H. Schwab, Vakuumisolationspaneele - Gas- und Feuchteeintrag sowie Feuchte-un Warmetransport, PhD thesis, Bayerischen Julius-Maximilians-UniversitatWurzburg, Wurzburg, 2004, 115 pp.

[3] D. Rath, Warmetransport in evakuierten Dammsystemen, PhD, UniversitatDortmund, Dortmund, 1989.

[4] K.H. Brodt, Thermal Insulations: Cfc-alternatives and Vacuum Insulation, PhDthesis, Delft University of Technology, Delft, 1995.

[5] J. Fricke, From Dewars to VIPs - one century of progress in vacuum insulationtechnology, in: M. Zimmerman (Ed.), Proceedings of the 7th InternationalVacuum Insulation Symposium, EMPA, Duebendorf, Switzerland, 28–29 Sep-tember 2005, (2005), pp. 5–14.

[6] M. Wutz, H. Adam, W. Walcher, Theorie und Praxis der Vakuumtechniek, ISBN 3-528-14884-5. Verlag Fried, Wiesbaden. Vieweg & Sohn, 2000.

[7] E.H. Kennard, Kinetic theory of Gases, with an Introduction to StatisticalMechanics, Mc-Graw-Hill, New York, 1938.

[8] R. Caps, U. Heinemann, M. Ehrmanntraut, J. Fricke, Evacuated insulation panelswith pyrogenic silica powders: properties and applications, High Temperatures-High Pressures 33 (2001) 151–156.

[9] J.-S. Kwon, C.H. Jang, H. Jung, T.-H. Song, Effective thermal conductivity of variousfilling materials for vacuum insulation panels, International Journal of Heat andMass Transfer 52 (2009) 5525–5532.

[10] B. Morel, D. Autissier, L. Autissier, Modifications of pyrogenic silica exposed tomoist air, in: A. Beck, et al. (Eds.), Proceedings of the 8th International VacuumInsulation Symposium, ZAE Bayern/UniWue, Wurzburg, September 18–19,2007, (2007), pp. 1–8.

[11] H. Schwab, U. Heinemann, J. Wachtel, H.P. Ebert, J. Fricke, Prediction for theincrease in pressure and water content of vacuum insulation panels (VIPs)integrated into building constructions using model calculations, Journal ofThermal Envelope and Building Science 28 (2005) 327–345.

[12] R. Caps, J. Fricke, Thermal conductivity of opacified powder filler materials forvacuum insulation, International Journal of Thermophysics 200 21 (2) (2000)445–452.

[13] S.R. Hostler, A.R. Abramson, M.D. Gawryla, S.A. Bandi, D.A. Schiraldi, Thermalconductivity of clay-based aerogel, International Journal of Heat and MassTransfer (2008), doi:10.1016/j.ijheatmasstransfer.2008.07.002.

[14] M. Reim, W. Korner, J. Manara, S. Korder, M. Arduini-Schuster, H.-P. Ebert, J.Fricke, Silica aerogel granulate material for highly thermal insulation and day-lighting, Solar Energy 79 (2005) 131–139.

[15] J. Fricke, H. Schwab, U. Heinemann, Vacuum insulation panels—exciting thermalproperties and most challenging applications, International Journal of Thermo-physics 27 (4) (2006) 1123–1139.

[16] H.C. Hottel, A.I. Sarofilm, Radiative Transfer, McGraw-Hill, 1967.[17] U. Heinemann, Influence of water on the total heat transfer in ‘evacuated’

insulations, in: Proceedings of the 7th International Vacuum Insulation Sympo-sium, 2005, pp. 23–34.

[18] A. Beck, F. Oswald, M. Binder, Influence of water content on the thermalconductivity of vacuum panels with fumed silica kernels, in: A. Beck, et al.(Eds.), Proceedings of the 8th International Vacuum Insulation Symposium, ZAEBayern/UniWue, Wurzburg, September 18–19, 2007, (2007), pp. 1–9.

[19] H. Schwab, U. Heinemann, A. Beck, H.P. Ebert, J. Fricke, Dependence of thermalconductivity on water content in vacuum insulation panels with fumed silicakernels, Journal of Thermal Envelope and Building Science 28 (2005) 319–326.

[20] U. Heinemann, Influence of water on the total heat transfer in ‘evacuated’insulations, International Journal of Thermophysics 29 (2008) 735–749.

http://dx.doi.org/10.1016/j.ijheatmasstransfer.2008.07.002


[21] R. Coquard, D. Quenard, Modelling of heat transfer in nanoporous silica—influ-ence of moisture, in: A. Beck, et al. (Eds.), Proceedings of the 8th InternationalVacuum Insulation Symposium, ZAE Bayern/UniWue, Wurzburg, September 18–19, 2007, (2007), pp. 1–13.

[22] W. Platzer, Optimisation and testing of a VIP exterior thermal insulation com-posite system (ETICS), in: A. Beck, et al. (Eds.), Proceedings of the 8th Interna-tional Vacuum Insulation Symposium, ZAE Bayern/UniWue, Wurzburg,September 18–19, 2007, (2007), pp. 1–8.

[23] H. Schwab, U. Heinemann, A. Beck, H.P. Ebert, J. Fricke, Permeation of differentgases through foils used as envelopes for vacuum insulation panels, Journal ofThermal Envelope and Building Science 28 (2005) 293–318.

[24] S. Brunner, Ph. Gasser, H. Simmler, K. Ghazi, Investigation of multilayeredaluminium-coated polymer laminates by focused ion beam (FIB) etching, Sur-face & Coatings Technology 200 (2006) 5908–5914.

[25] S. Brunner, P.J. Tharian, H. Simmler, K. Ghazi, Focused ion beam (FIB) etching toinvestigate aluminium-coated polymer laminates subjected to heat and moist-ure loads, Surface & Coatings Technology 202 (2008) 6054–6063.

[26] R. Caps, H. Beyrichen, D. Kraus, S. Weismann, Quality control of vacuuminsulation panels: methods of measuring gas pressure, Vacuum 82 (2008)691–699.

[27] H. Schwab, C. Stark, J. Watchel, H.P. Ebert, J. Fricke, Thermal bridges in vacuum-insulated building facades, Journal of Thermal Envelope and Building Science 28(2005) 345–355.

[28] K. Ghazi, R. Bundi, B. Binder, Effective thermal conductivity of vacuum insulationpanels, Building Research and Information 32 (4) (2004) 293–299.

[29] M.J. Tenpierik, J.J.M. Cauberg, Analytical models for calculating thermal bridgeeffects caused by thin high barrier envelopes around vacuum insulation panels,Journal of Building Physics 30 (2007) 185–215.

[30] M.J. Tenpierik, W. van der Spoel, J.J.M. Cauberg, Analytical models for calculatingthermal bridge effects in high performance building enclosure, Journal ofBuilding Physics 31 (2008) 361–388.

[31] M.K. Willems, K. Schild, G. Hellinger, Numerical investigation on thermal bridgeeffects in vacuum insulating elements, in: M. Zimmerman (Ed.), Proceedings ofthe 7th International Vacuum Insulation Symposium, EMPA, Duebendorf, Swit-zerland, 28–29 September 2005, (2005), pp. 145–152.

[32] D. Quenard, H. Sallee, From VIP’s to building facades: three levels of thermalbridges, in: M. Zimmerman (Ed.), Proceedings of the 7th International VacuumInsulation Symposium, EMPA, Duebendorf, Switzerland, 28–29 September 2005,(2005), pp. 113–120.

[33] R. Bundi, K. Ghazi, Th. Frank, Vacuum insulated panels in building applications,in: Int. Conference, CISBAT, EPFL, Lausanne, 2003.

[34] K. Ghazi, T. Nussbaumer, R. Bundi, Thermal performance of VIP assemblies inbuilding constructions, in: M. Zimmerman (Ed.), Proceedings of the 7th Inter-national Vacuum Insulation Symposium, EMPA, Duebendorf, Switzerland, 28–29September 2005, (2005), pp. 131–138.

[35] H.A. Fine, Advanced evacuated thermal insulations: the state of the art, Journal ofBuilding Physics 12 (3) (1989) 183–208.

[36] L.R. Glicksman, Two-dimensional heat transfer effects on vacuum and reflectiveinsulations, Journal of Building Physics 14 (4) (1991) 281–294.

[37] T.G. Kollie, D.L. McElroy, H.A. Fine, K.W. Childs, R.S. Graves, F.J. Weaver, A Reviewof Vacuum Insulation Research and Development in the Building MaterialsGroup of the Oak Ridge National Laboratory. ORNL/TM-11703, 1991.

[38] K.G. Degen, S. Rosetto, T. Reichenauer, J. Fricke, Investigations of heat transfer inevacuated foil—spacer multilayer insulations, Journal of Building Physics 16 (2)(1992) 140–152.

[39] B.T. Griffith, D. Arasteh, Advanced insulations for refrigerators/freezers—thepotential for new shell designs incorporating polymer barrier construction,Energy and Buildings 22 (3) (1995) 219–231.

[40] P. Mukhopadhyaya, M.K. Kumaran, N. Normandin, D. van Reenen, J.C. Lackey,High performance vacuum insulation panel: development of alternative corematerials, Journal of Cold Regions Engineering 22 (4) (2008) 103–123.

[41] L.R. Glicksman, M.S. Burke, US Patent 5,032,439—thermal insulations usingvacuum panels, 1991.

[42] K. van Went, Vacuum Isolatie Panelen, Kwalificering en kwantificering van dethermische eigenschappen van vacuum isolatie panelen en mogelijke toepas-singen in de bouw, M.Sc. thesis, 2002.

[43] T. Nussbaumer, K. Ghazi, Ch. Tanner, Experimental and numerical investigationof the thermal performance of a protected vacuum insulation system applied to aconcrete wall, Applied Energy 83 (8) (2006) 841–855.

[44] T.I. Thorsell, I. Kallebrink, Edge loss minimization in vacuum insulation panels,in: Proceedings of the 7th Symposium on Building Physics in the NordicCountries, 13–15 June 2005, The Icelandic Building Research Institute, Reykjavik,2005, pp. 945–952.

[45] T. Nussbaumer, R. Bundi, Ch. Tanner, H. Muehlebach, Thermal analysis of awooden door system with integrated vacuum insulation panels, Energy andBuildings 37 (2005) 1107–1113.

[46] J.J.M. Cauberg, M.J. Tenpierik, From VIP to building panel, in: M. Zimmerman(Ed.), Proceedings of the 7th International Vacuum Insulation Symposium,EMPA, Duebendorf, Switzerland, 28–29 September 2005, (2005), pp. 161–171.

[47] Lacasse, Sjostrom, Recent advances in methods for service life prediction ofbuilding materials and components—an overview, in: Proceedings of the CIBWorld Building Congress, Toronto, Ontario, May 2, 2004, (2004), pp. 1–10.

[48] H. Simmler, S. Brunner, Ageing and service life of VIP in buildings, in: M.Zimmerman (Ed.), Proceedings of the 7th International Vacuum InsulationSymposium, EMPA, Duebendorf, Switzerland, 28–29 September 2005, (2005),pp. 15–22.

[49] M.J. Tenpierik, W. van der Spoel, J.J.M. Cauberg, Simplified analytical models forservice life prediction of a vacuum insulation panel, in: A. Beck, et al. (Eds.),Proceedings of the 8th International Vacuum Insulation Symposium, ZAEBAyern/UniWue, Wurzburg, September 18–19, 2007, (2007), pp. 1–8.

[50] M.J. Tenpierik, J.J.M. Cauberg, Vacuum insulation panels in building facades:moisture and temperature conditions during installation, in: Proceedings of the7th Symposium on Building Physics in the Nordic Countries, IBRI, Reykjavik, June13–15, 2005, (2005), pp. 937–944.

[51] H. Schwab, U. Heinemann, A. Beck, H.P. Ebert, J. Fricke, Prediction of service lifefor vacuum insulation panels with fumed silica kernel and foil cover, Journal ofThermal Envelope and Building Science 28 (2005) 357–375.

[52] H. Simmler, S. Brunner, Vacuum insulation panels for buildings—basic proper-ties, ageing mechanisms and service life, Energy and Buildings 37 (2005)1122–1131.

[53] S. Brunner, H. Simmler, In situ performance assessment of vacuum insulationpanels in a flat roof construction, Vacuum 82 (2008) 700–707.

[54] R. Caps, Monitoring gas pressure in vacuum insulation panels, in: M. Zimmer-man (Ed.), Proceedings of the 7th International Vacuum Insulation Symposium,EMPA, Duebendorf, Switzerland, September 28–29, (2005), pp. 57–66.

[55] A.Z. Uzzo, Tire Conditioning Monitor System, United States Patent 4 609 905(1986).

[56] B.P. Wilson, R.J. Trew, R.W. Koch, Combination monitoring device and patch for apneumatic tire and method of installing the same with a coupled antenna,United States Patent 6 208 244 (2001).

[57] B.P. Jelle, T.-N. Nilsen, P.J. Hovde, A. Gustavsen, Accelerated climate ageing ofbuilding materials and application of the attenuated total reflectance (ATR)Fourier transformed infrared (FTIR) radiation experimental method, in: C. Rode(Ed.), Proceedings of the 8th Symposium on Building Physics in the NordicCountries, vol. 2, Copenhagen, Denmark, June 16–18, (2008), pp. 951–958.

[58] T.I. Thorsell, Vacuum Insulation in Buildings—Means to Prolong the Service Life,Licentiate thesis, KTH-The Royal Institute of Technology, Stockholm, Sweden,2006.

[59] B.P. Jelle, A. Gustavsen, T.N. Nilsen, T. Jacobsen, Solar material protection factor(SMPF) and solar skin protection factor (SSPF) for window panes and other glassstructures in buildings, Solar Energy materials and Solar Cells 91 (2007) 342–352.

[60] K. Lenz, Phl. Leistner, K. Sedlbauer, N. Konig, Vakuumisolationspaneele aushygrothermische und akustischer Sicht, in: VIP-Bau (Ed.), Proceedings of the2nd Fachtagung ‘‘Erfahrungen aus der Praxis’’, Wismar, Germany, 16–17 June,2005, E1–E13, 2005.

[61] J.J.M. Cauberg, M.J. Tenpierik, Sound reduction of vacuum insulation basedsandwich panels, in: A. Beck, et al. (Eds.), Proceedings of the 8th InternationalVacuum Insulation Symposium, ZAE Bayern/UniWue, Wurzburg, September 18–19, 2007, (2007), pp. 1–8.

[62] W. Maysenholder, Sound Transmission Loss of Vacuum Insulation Panels,Fraunhofer Institute for Building Physics, Stuttgart, Germany, 2008.

[63] J. Cremers, Einsatzmoglichkeiten von Vakuum-Dammsystemen im Bereich derGebaudehulle, PhD thesis, Technische Universitat Munchen, Munchen, 2006,248 pp.

[64] H. Schwab, J. Wachtel, H. Scheuerpflug, C. Stark, U. Heinemann, H.P. Ebert, J.Fricke, Entwicklung und Anwendung von evakuierten hocheffizienten Dam-mungen fur Gebaude (Vakuumdammung fur Gebaude), Bayerische Zentrumfur Angewandte Energieforschung E.V., Bayern, 2003.

[65] W.M. Willems, K. Schild, Vacuum insulated constructions in detail, in: Proceed-ings of the 7th Symposium on Building Physics in the Nordic Countries, IBRI,Reykjavik, June 13–15, 2005, (2005), pp. 928–936.

[66] H. Winkler, G.-W. Mainka, Development and first experiences of a prefabricatedVIP-sandwich element for fast and secure application on building surfaces, in: M.Zimmerman (Ed.), Proceedings of the 7th International Vacuum InsulationSymposium, EMPA, Duebendorf, Switzerland, 28–29 September 2005, (2005),pp. 121–130.

[67] M. Zwerger, H. Klein, Integration of VIP’s into external wall insulation systems,in: M. Zimmerman (Ed.), Proceedings of the 7th International Vacuum InsulationSymposium, EMPA, Duebendorf, Switzerland, 28–29 September 2005, (2005),pp. 173–179.

[68] Z.A.E. Bayern, Abschlussbericht VIP - Demoobjekte, Bayerische Zentrum furAngewandte Energie-forschung E.V., Bayern, 2008, Retrieved October 9, 2008,from www.zae-beyern.de.

[69] A. Binz, G. Steinke, Applications of vacuum insulation in the building sector, in:M. Zimmerman (Ed.), Proceedings of the 7th International Vacuum InsulationSymposium, EMPA, Duebendorf, Switzerland, 28–29 September 2005, (2005),pp. 43–48.

[70] R. Ogden, C. Kendrick, VIP cladding panels for buildings: applications andconceptual solutions, in: M. Zimmerman (Ed.), Proceedings of the 7th Interna-tional Vacuum Insulation Symposium, EMPA, Duebendorf, Switzerland, 28–29September 2005, (2005), pp. 153–160.

[71] M.J. Tenpierik, J.J.M. Cauberg, VIP integrated facade designs: the advantage ofcombining high thermal performance with limited construction thickness, in:B.K. Tan, S.K. Wittkopf (Eds.), Proceedings of the 24th International Conferenceon Passive and Low Energy Architecture, PLEA/NUS, Singapore, September 22–24, 2007, (2007), pp. 303–310.

[72] J.J.M. Cauberg, M.J. Tenpierik, A 4.1 Vacuumisolatiepanelen en andere noviteiten,in: D.W. Dicke, E.M. Haas (Eds.), Praktijkhandboek Duurzaam Bouwen, WekaPublishers, Amsterdam, Netherlands, 2005, , 4.1 VAC 1-20.

[73] S. Brunner, H. Simmler, In situ performance assessment and service life ofvacuum insulation panels (VIP) in buildings, in: A. Beck, et al. (Eds.), Proceedings

http://www.zae-beyern.de/


of the 8th International Vacuum Insulation Symposium, ZAE Bayern/UniWue,Wurzburg, September 18–19, 2007, (2007), pp. 1–8.

[74] M. Jordi, Development of a Low Energy House with Vacuum Insulation, RenggliAG Suree and Schotz, Switzerland, 2008, Retrieved November 19, 2008 fromhttp://www.vip-bau.ch/.

[75] J.H. Kerspe, K. Zwarts, Newly designed vacuum insulation for big pipes andpipelines, in: A. Beck, et al. (Eds.), Proceedings of the 8th International VacuumInsulation Symposium, ZAE Bayern/UniWue, Wurzburg, September 18–19,2007, (2007), pp. 1–13.

[76] S.A. Omer, S.B. Riffat, G. Qiu, Technical note: thermal insulation for hot watercylinders: a review and a conceptual evaluation, Building Services EngineeringResearch and Technology 28 (3) (2007) 275–293.

[77] Cauberg-Huygen Raadgevende Ingenieurs B.V., Toepassingsmogelijkheden vanVacuum Isolatie Panelen bij LTV vloer- en wandverwarmingssystemen inwoningen, 2004, Retrieved October 3, 2008, from Tenpierik, J.M.

[78] Cauberg-Huygen Raadgevende Ingenieurs B.V., Toepassingsmogelijkheden vanVacuumisolatie-panelen in gebouwinstallaties, 2006, Retrieved October 3, 2008,from Tenpierik, J.M.

[79] B.T. Griffith, D. Arashteh, Gas-filled panels: a thermally improved buildinginsulation. Thermal performance of the exterior envelopes of buildings V, in:Proceedings of the ASHRAE/DOE/BTECC Conference, Clear Water Beach, Florida,7–10 December 1992, (1992), pp. 1–13.

[80] B.T. Griffith, D. Arashteh, D. Turler, Gas-filled panels: an update on applicationsin the building thermal envelope, in: Superinsulations and the Building Envel-ope, Proceedings of the BETEC Fall Symposium, Washington, District of Colum-bia, 14 November 1995, (1995), pp. 1–13.

[81] S.Q. Zeng, A.J. Hunt, W. Cao, R. GReif, Pore size distribution and apparent thermalconductivity of silica aerogel, Journal of Heat Transfer 116 (1994) 756–759.

[82] M.J. van Bommel, C.W. den Engelsen, J.C. van Miltenburg, A thermoporometrystudy of fumed silica/aerogel composites, Journal of Porous Materials 4 (1997)143–150.

[83] I. Smirnova, Synthesis of Silica Aerogels and their Applications as a Drug DeliverySystem, PhD thesis, Technischen Universitait berlin, Germany, 2002, 154 pp.

[84] K. Richter, P.M. Norris, C.-L. Chang, Aerogels: applications, structure and heattransfer phenomena, in: V. Prasad, Y. Jaluria, G. Chen (Eds.), Annual Review onHeat Transfer, vol. 6, 1995, 61–114.

[85] N. Husing, U. Schubert, Aerogele—luftige Materialien: Chemie, Struktur undEigenschaften,(Aerogels—airy materials: chemistry, structure and properties),Angewandte Chemie 110 (1998) 22–47, http://gfp.lbl.gov/, Retrieved 2008.

[86] D.M. Smith, A. Maskara, U. Boes, Section 7. Properties. A. Thermal—aerogel-basedthermal insulation, Journal of Non-Crystalline Solids 225 (1997) 254–259.

[87] K. Ramakrishnan, A. Krishnan, V. Shankar, I. Srivastava, A. Singh, R. Radha,Modern Aerogels, 2007, Retrieved September 27, 2008, from www.dstuns.iitm.ac.in.

[88] L. Pei, K.-I. Kurumada, M. Tanigaki, M. Hiro, K. Susa, Closed-porosity mesoporoussilica produced by high temperature rapid drying, Journal of Materials Science39 (2004) 663–665.

[89] L. Pei, K.-I. Kurumada, M. Tanigaki, M. Hiro, K. Susa, Effect of drying on themesoporous structure of sol–gel derived silica with PPO-PEO-PPO templateblock copolymer, Journal of Colloid and Interface Science 284 (2005) 222–227.

[90] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T.W.Chu, D.H. Olsen, E.W. Sheppard, S.B. McCullen, J.B. Higgins, J.L. Schlenker, A newfamily of mesoporous molecular sieves prepared with liquid crystal templates,Journal of the American Chemical Society 144 (27) (1992) 10834–10843.

[91] D. Zhao, Q. Huo, J. Feng, B.F. Chmelka, G.D. Stucky, Nonionic triblock and stardiblock copolymer and oligomeric surfactant syntheses of highly ordered,hydrothermally stable, mesoporous silica structures, Journal of the AmericanChemical Society 120 (24) (1998) 6024–6036.

[92] L.F. Giraldo, L. Lopez, L. Perez, S. Urrego, L. Sierra, M. Mesa, Mesoporous silicaapplications, Macromolecular Symposia 258 (2007) 129–141.

[93] B.P. Jelle, G. Hagen, Performance of an electrochromic window based on polyani-line, prussian blue and tungsten oxide, Solar Energy Materials and Solar Cells 58(1999) 277–286.

[94] B.P. Jelle, Electrochemical and Spectroscopic Studies of Electrochromic Materi-als, PhD thesis, Norges Tekniske Høgskole, Trondheim, Norway, 1993.

[95] M. Cohen, J. Heijnis, The future envelope, towards a more reactive facade, in:Proceedings of the Future Envelope. Conference on Climate-oriented BuildingEnvelopes, Delft, The Netherlands, June 15, 2008, (2007), pp. 1–5.

[96] R. Baetens, Properties, Requirements and Possibilities for Highly Thermal Insu-lating Materials and Solutions in Buildings—State-of-The-Art and Beyond, M.Sc.thesis, Catholic University Leuven, Leuven, Belgium—SINTEF Building and Infra-structure, Trondheim, Norway—Norwegian University of Science and Technol-ogy, Trondheim, Norway, 2009.

[97] A. Binz, M. Erb, Hp. Eicher, H. Simmler, W.K. Ghazi, Vakuum-Dammung imbaubereich vip-bau.ch, 2005, Retrieved November 19, 2008 from http://www.empa.ch/.

[98] U. Heineman, H. Schwab, H. Simmler, S. Brunner, K. Ghazi, R. Bundi, K. Kumaran,Ph. Mukhopadhyaya, D. Quenard, H. Sallee, K. Noller, E. Kucukpinar-Niarchos, C.Stramm, M.J. Tenpierik, J.J.M. Cauberg, A. Binz, G. Steinke, A. Moosmann, M. Erb,Vacuum Insulation. Panel Properties and Building Applications (Summary), FinalReport for the IEA/ECBCS Annex 39 HiPTI-project (High Performance ThermalInsulation for Buildings and Building Systems), 2005.

[99] H. Simmler, S. Brunner, U. Heinemann, H. Schwab, K. Kumaran, Ph. Mukhopad-hyaya, D. Quenard, H. Sallee, K. Noller, E. Kucukpinar-Niarchos, C. Stramm, M.J.Tenpierik, J.J.M. Cauberg, M. Erb, Vacuum Insulation Panels. Study on VIP-components and Panels for Service Life Prediction in Building Applications(Subtask A), Final Report for the IEA/ECBCS Annex 39 HiPTI-project (HighPerformance Thermal Insulation for Buildings and Building Systems), 2005a.

[100] A. Binz, A. Moosmann, G. Steinke, U. Schonhardt, F. Fregnan, H. Simmler, S.Brunner, K. Ghazi, R. Bundi, U. Heinemann, H. Schwab, J.J.M. Cauberg, M.J.Tenpierik, G. Johannesson, T. Thorsell, M. Erb, B. Nussbaumer, Vacuum Insula-tion in the Building Sector. Systems and Applications (Subtask B), Final Reportfor the IEA/ECBCS Annex 39 HiPTI-project (High Performance Thermal Insulationfor Buildings and Building Systems), 2005b.

[101] International Standard ISO 15686-2: 2001. (E). Buildings and ConstructedAssets—Service Life Planning—Part 2: Service Life Prediction Procedures.

[102] K.I. Jensen, J.M. Schultz, Transparent Aerogel Windows—results from an EU FP5project, in: A. Beck, et al. (Eds.), Proceedings of the 8th International VacuumInsulation Symposium, ZAE Bayern/UniWue, Wurzburg, September 18–19,2007, (2007), pp. 1–8.

[103] R. Caps, J. Hetfleisch, F. Jochen. Offenlegungsschrift DE 196 47 567 A 1. GermanPatent (1998).

[105] J. Cremers, Typology of applications for opaque and translucent VIP in thebuilding envelope and their potential for temporary thermal insulation, in:Presentation for the 7th International Vacuum Insulation Symposium, EMPA,Duebendorf, Switzerland, September 28–29, (2005), pp. 57–66.

[106] W.M. Willems, K. Schild, The use of Vacuum Insulated Sandwiches (VIS)in building envelopes, in: Proceedings of the 4th European Conference on EnergyPerformance & Indoor Climate in Buildings, Portland, Lyon, France, (2006), pp.1–6.

[107] R. Bundi, Faschung und Entwicklung im Fassadenbau, Vakuumisolierte Paneele,Fassade 3 (2003) 19–22.

[108] M.J. Tenpierik, J.J.M. Cauberg, T.I. Thorsell, Integrating vacuum insulation panelsin building constructions: an integral perspective, Construction Innovation 7 (1)(2007) 38–53.

[109] B.T. Griffith, D. Turler, D. Arashteh, Optimizing the effective conductivity andcost of gas-filled panel thermal insulations, in: Proceedings of the 22nd Inter-national Thermal Conductivity Conference, Arizona State University, 7–10November 1993, (1993), pp. 1–12.

[110] K. Kuroda, Silica-Based Mesoporous Molecular Sieves Derived from a LayeredPolysilicate Kanemte - A Review, Journal of Porous Materials 3 (1996) 107–114.

http://www.vip-bau.ch/

http://gfp.lbl.gov/

http://www.dstuns.iitm.ac.in/

http://www.dstuns.iitm.ac.in/

http://www.empa.ch/

http://www.empa.ch/

Vacuum insulation panels for building applications: A review and ...

Documents

Transcript of Vacuum insulation panels for building applications: A review and ...