hygrothermal performance of an experimental hemp-lime building
TRANSCRIPT
Construction and Building Materials 36 (2012) 270–275
Contents lists available at SciVerse ScienceDirect
Construction and Building Materials
journal homepage: www.elsevier .com/locate /conbui ldmat
Hygrothermal performance of an experimental hemp–lime building
Andy Shea ⇑, Mike Lawrence 1, Pete Walker 2
BRE Centre for Innovative Construction Materials, Department of Architecture and Civil Engineering, University of Bath, Bath, BA2 7AY, UK
h i g h l i g h t s
" We review the environmental benefits of building with Hemp–Lime." We report the construction of a full-scale Hemp–Lime test building." We present the thermal performance testing of the completed building." A total heat loss coefficient for the test building was determined as 36.7 W/K." Laboratory tests show the time taken to reach a steady-state condition is in the order of 10 days.
a r t i c l e i n f o
Article history:Received 13 February 2012Received in revised form 16 April 2012Accepted 29 April 2012Available online 23 June 2012
Keywords:Hemp–limeHygrothermal performanceCo-heatingPhase-changeDampingLow-carbon building design
0950-0618/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.conbuildmat.2012.04.123
⇑ Corresponding author. Tel.: +44 (0) 1225 386158E-mail addresses: [email protected] (A. Sh
(M. Lawrence), [email protected] (P. Walker).1 Tel.: +44 (0) 1579 345355.2 Tel.: +44 (0) 1225 386646.
a b s t r a c t
The use of hemp–lime as a construction material combines renewable low carbon materials with excep-tional hygrothermal performance. The hemp plant can grow up to 4 m in 4 months, with low fertilizerand irrigation demand, making it very efficient in the use of time and material resources. All parts ofthe plant can be used – the seed for food stuffs, the fibre surrounding the stem for paper, clothing andresin reinforcement, and the woody core of the stem as animal bedding and aggregate in hemp–lime con-struction. The unique pore structure of the woody core (shiv) confers relatively low thermal conductivityand hygric buffering. The construction technique promotes good air tightness and minimal thermalbridging of the building envelope. All these factors combine to produce low carbon, hygrothermally effi-cient buildings that are low energy both in construction and in use, and offer opportunities for recyclingat end of life. This paper presents the hygrothermal performance of an experimental hemp–lime buildingand compares the results of steady-state co-heating tests with laboratory tests and computer simulationsof transient performance.
� 2012 Elsevier Ltd. All rights reserved.
1. Introduction
Carbon dioxide (CO2) is a greenhouse gas which is considered tobe the major contributory factor in global warming. As a result,international efforts are being made to reduce greenhouse gasemissions, including CO2, to below 1990 levels. In 2008 emissionsof CO2 in the UK in were 525 million tonnes (Mt) compared with590 Mt in 1990 [1]. Of this figure, the construction sector wasresponsible for 298.4 Mt [2]. Table 1 presents a breakdown of theCO2 emissions contribution of the construction industry as partof total UK carbon emissions.
The construction sector has a greater influence on carbon emis-sions (56.7%) than all other sectors combined. Within buildings theuse of services such as heating, lighting and air conditioning is
ll rights reserved.
.ea), [email protected]
responsible for nearly 47% of CO2 emissions in the UK, whilst man-ufacture of building materials is responsible for nearly 9%. It iswithin these two areas that the focus on the reduction in CO2 emis-sions (carbon reduction) has concentrated. Carbon reduction in useis associated with improvements in thermal insulation, increasedefficiency of lighting, heating and cooling and reduction in thermallosses through thermal bridges and poor air-tightness. Carbonreduction in manufacturing is associated with a reduction inenergy input in the manufacturing process (low carbon cements,substitution of high carbon materials with lower carbon ones).Another area of interest is the use of building materials thatsequester CO2. This is most often achieved through the use of nat-ural plant-fibre building materials which absorb atmospheric CO2
through photosynthesis thereby locking it up within the materialfor the lifetime of the building. Many of these natural plant-fibrematerials offer opportunities for recycling and re-use, which fur-ther extends their useful life. Some materials can be compostedand recycled through agriculture. Accordingly, interest in materialssuch as timber, straw and hemp has grown in recent years in
Table 1Construction industry CO2 emissions in 2008.
Sub-sector: Construction CO2 (Mt) % of total
Design 1.3 0.25Manufacture 45.2 8.61Distribution 2.8 0.53Operations on-site 2.6 0.50In Use 246.4 46.93Refurb/demolition 1.3 0.25Total construction 298.4 56.84Other sectors 226.6 43.16Total UK 525.0 100
A. Shea et al. / Construction and Building Materials 36 (2012) 270–275 271
response to the need to source materials with a lower environmen-tal impact.
Table 2Mechanical and thermal characteristics of hemp–lime [4].
Application Shiv – binderproportions(by mass)
Density(kg/m3)
Compressivestrength (N/mm2)
Thermalconductivity(W/m K)
Roof insulation 1:1 220 0.05 0.06Wall construction 1:1.5 275 0.11 0.06–0.09Wall construction 1:2 330 0.22 0.09–0.115Wall construction 1:2
(compressed)440 0.35 0.115
Floor 1:3 500 0.8 0.13Floor 1:4 600 1.15 0.14Pre-cast
structural1:4(compressed)
600–1000
2–6 0.14–0.27
1.1. Hemp–lime
Hemp–lime is a building material often referred to in English asHemcrete� or Lime–Hemp. This material was originally developedas a replacement for wattle and daub infill in timber frame build-ings in France, where the term used is Chaux–Chanvre. It is madeby mixing the chopped woody core of the stalks of the hemp plant(cannabis sativa), known as the ‘shiv’, with a binder made from airlime with pozzolanic, cementitious or hydraulic lime additions,and in some cases small amounts of other additives such as surfac-tants. The material is used to form building envelopes by castingbetween, or spraying against, temporary or permanent shutteringin situ, or by pre-fabrication of building blocks or panels. Typicallywalls are constructed to be �300 mm in thickness. Hemp shiv canalso be used as an insulating element in lime renders. Interest inthe use of hemp–lime is driven by the following factors [3,4]:
� It is a low density material with associated low thermalconductivity.� Its pore structure allows it to dampen variations in environ-
mental heat and humidity.� The high proportion of embodied bio-based material results in
the sequestration of relatively large amounts of atmosphericCO2 (through photosynthesis), compared with more traditionalbuilding materials.� Hemp shiv is more resistant to biological decay than some other
bio-based building materials (for example straw);� Hemp shiv, in common with other bio-based materials, is a
renewable resource, and also offers the opportunity of beingrecycled at end of life.� Hemp cultivation requires lower levels of fertilisation and irri-
gation than wheat and some other bio-based building materials,resulting in lower levels of eutrophication.� The hemp plant grows very rapidly to heights of up to 4 m
within 4 months. This gives it the potential to act as a ‘breakcrop’ allowing optimisation of yields of the primary crop.
The density, thermal conductivity, and compressive strength ofhemp–lime are predominantly controlled by the relative propor-tions of shiv and binder. These characteristics are listed in Table2. Hemp shiv sequesters 2.1 kg CO2 equivalent per kg, and a 1 m2
timber-framed lime-rendered 300 mm thick wall made with a1:2 mix sequesters 75.7 kg CO2 equivalent with the net CO2 emis-sions including transport, construction and manufacturing pro-cesses (carbon footprint) being �35.5 kg CO2 equivalent [3],which equates to a negative carbon contribution by the wall ele-ment to the total carbon footprint of the construction.
A major advantage of natural fibre insulation materials is theirability to create a breathable wall construction by readily absorb-
ing and releasing moisture in response to changes in relativehumidity and vapour pressure gradients in the surrounding envi-ronment. Heat flows are associated with these reactions, duringabsorption heat is released and on release of moisture heat is ab-sorbed [5]. Previous research has presented the physical propertiesof hemp–lime [6–8] and highlighted that the material presents agood balance between low mass and heat storage capacity com-pared with classical insulation materials.
At present inorganic insulation materials dominate the buildingindustry, although interest in the use of natural fibre insulationproducts is steadily increasing [5]. In Europe inorganic fibrousmaterials, e.g. stone wool and glass wool, account for 60% of themarket. Organic foamy materials such as expanded and extrudedpolystyrene account for 27% of the market, whilst all other materi-als combined make up less than 13% [9].
Laboratory experiments and field tests conducted at the Univer-sity of Bath aimed to investigate the hygrothermal performance ofhemp–lime and to establish robust data on its use as a constructionmaterial for use in low-carbon buildings. Research and develop-ment has focused on the construction of a full-scale test building.The construction and performance of this building is reported here.The results of laboratory tests performed on a 300 mm hemp–limepanel, heat flow meter tests, and WUFI simulations are alsopresented.
2. Method
2.1. Hempod test building
An experimental full-scale test-building, referred to as the Hempod, was con-structed at The University of Bath, UK (Fig. 1). Construction of the Hempod beganin June 2010; taking 10 days to complete. The walls were left to dry for 8 weeks be-fore the external render was applied. Over the autumn period the building was al-lowed to dry out. During this period the moisture content of the walls wasmonitored and by January 2011 moisture levels inside the building had equilibratedwith the external environment. Accordingly, all experiments characterising the per-formance of the finished building were conducted after January 2011.
The single-zone building was constructed on a suspended chipboard floor insu-lated with 200 mm closed-cell insulation (k = 0.023 W/m K) and has a floor area ofapproximately 27 m2. The ceiling was also constructed of 200 mm closed-cell insu-lation behind a 9 mm layer of gypsum plasterboard. Conventional surface heattransfer coefficients for floors and ceilings were used and overall thermal transmit-tance was calculated as 0.14 W/m2 K for both the floor and the roof. Windows anddoors are timber-framed with low-emissivity triple-glazed argon-filled glazing. Thedoor has a U-value of 0.79 W/m2 K (south facing), the windows 0.97 W/m2 K (northfacing) and 1.05 W/m2 K (south facing). Junctions between wall and floor, wall andceiling, wall and door/windows were sealed with vapour permeable tape. The ceil-ing was lined underneath the plasterboard with Intello� vapour check membrane,sealed to the walls. Walls were formed from 75 � 50 mm timber studwork at600 mm centres to act as structural support. These were positioned on the interiorof the walls and clad with a permanent shuttering made from 9 mm thick magne-sium silicate board. A 200 mm thick hemp–lime wall was cast using temporaryshuttering, rising to above the level of the insulated ceiling. The mix used was 1 partTradical�HF hemp shiv to 1.5 parts Tradical�HB binder, with minimal compressionapplied in order to achieve a target density of 275 kg/m3.
Fig. 1. The completed HemPod building at The University of Bath.
272 A. Shea et al. / Construction and Building Materials 36 (2012) 270–275
The apparent, i.e. measured, thermal conductivity of insulation materials alsoincludes the effects of other modes of heat transfer. Natural fibre and other insula-tion materials exhibit a ‘fish-hook’ measured conductivity for varying density,where at lower densities measured conductivity is high but there is a decrease withincreasing density until a minimum is reached from which point conductivity in-creases with increasing density. At low densities the impact of long-wave radiationis apparent resulting in increased conductivity, then, above the optimum density,increasing the density causes the thermal conductivity to also increase because ofincreased conduction through the material [10]. The range of densities at whichany practical building use of hemp–lime can be made is above the minimum con-ductivity value. Laboratory assessment using a heat flow meter compared a rangeof hemp–lime samples of differing densities and formulations at higher than targetdensity (Fig. 2) to evaluate the extent to which thermal conductivity increased withhigher density, and these data were used to estimate the conductivity of the hemp–lime panels of the Hempod in the field where the average water content of the pan-els was estimated to raise the density to approximately 310 kg/m3.
Based on a density of 310 kg/m3, conductivity of the hemp–lime was taken asapproximately 0.08 W/m K. The thermal conductivity of the magnesium silicateboard was 0.26 W/m K, and timber studwork was assumed to be 0.12 W/m K. Sur-face resistances were estimated using:
Rs ¼1
hc þ hrð1Þ
where hc is the convective coefficient; hr is the radiative coefficient; and
hr ¼ ehr0 ð2Þ
hr0 ¼ 4rT3m ð3Þ
where e is the hemispherical emissivity of the surface; hr0 is the radiative coefficientfor a black-body surface (W/m2 K); r is the Stefan–Boltzmann constant(5.67 � 10�8 W/m2 K4); Tm is the mean thermodynamic temperature of the surfaceand of its surroundings (K).
Fig. 2. Measured thermal conductivity of hemp–lime samples at a range ofdensities.
At the interior surface hc = hci and a typical value for horizontal heat flow is2.5 W/m2 K. Thus for Tm = 293 (20 �C) and surface emissivity 0.91, the interior heattransfer coefficient is 7.69 W/m2 K giving Rsi = 0.13 m2 K/W. At the exterior surface,hc = hce and:
hce ¼ 4þ 4v ð4Þ
where v is the wind speed adjacent to the surface (m/s).Assuming overcast sky conditions, Tm = 283 (10 �C), and a wind speed of 4 m/s,
the exterior heat transfer coefficient is 25.1 W/m2 K, giving Rse = 0.04 m2 K/W. Toaccount for the timber studwork within the depth of the hemp–lime, the total ther-mal resistance is calculated as the arithmetic mean of the upper and lower limits ofthe thermal resistance as described in BS EN ISO 6946 [11]. The total thermal resis-tance (environment to environment) of the wall element was calculated to be3.20 m2 K/W, corresponding to a thermal transmittance (U-value) of 0.42 W/m2 K.Using the calculated U-values and measured surface areas of the Hempod, an eval-uation of the steady-state thermal performance was conducted using a simplespread-sheet model, which was compared with the results of a co-heating test.The co-heating test used a single electric heater and circulation fan, linked to a ther-mostat, to provide a steady, above ambient, indoor air temperature. A detaileddescription of the co-heating test methodology is presented in Palmer et al. [12].The Hempod was heated to approximately 24 �C and the total energy input to thebuilding from the operation of the fan and electric heater was metered. 7 days’ datawere used for the calculation of the building heat loss coefficient3. Solar radiationwas measured using a thermopile-type pyranometer and regression analysis usedto remove the effect of heating due to solar gain. The resulting heat loss coefficientrepresents the combined fabric and infiltration losses from the finished buildingand was calculated as 36.7 W/K. The air permeability of the Hempod was tested priorto the co-heating test in accordance with the procedures detailed in ATTMA TS1 [13]and BS EN 13829 [14] test Method A, thus the building was tested in its finished statewith no temporary seals. The number of air changes per hour at 50 Pa (n50) was 0.55,which is within the PassivHaus limiting value of 60.6 air changes per hour. Heat lossdue to infiltration was incorporated into the simple steady-state model as was heatloss associated with thermal bridging; the latter being determined using default lin-ear thermal transmittance (W-values) as presented in Ward [15]. Assuming hemp–lime to have the k = 0.08 W/m K stated earlier, the steady-state model predictedwhole building heat loss as 35.6 W/K; within 3% of the co-heating test measurement.The percentage breakdown of these heat losses was 47.1% hemp-based fabric ele-ments, 34.3% non-hemp elements, 13.5% for thermal bridging, and 5.1% due toinfiltration.
On completion of the co-heating test, a thermographic study of the Hempodwas conducted to assess the effects of thermal bridging at openings and junctionswithin the building. Non-repeating thermal bridges lead to increased heat lossnot accounted for in the U-value calculation and result in reduced internal surfacetemperature, which leads to an increased risk of condensation and mould growth.BRE Information Paper 1/06 [15] presents a method for the assessment of the riskof surface condensation or mould growth for building details subject to specificinternal and external environmental conditions using a temperature factor fRsi. Tolimit the risk of surface condensation or mould growth, fRsi should be greater thanor equal to a critical value (fCRsi). fRsi is defined by:
fRsi ¼Tsi � Te
Ti þ Teð5Þ
where Tsi is the minimum internal surface temperature; Te is the external tempera-ture; Ti is the internal temperature.
The thermographic survey identified that the junction of the floor with thenorth and west facing walls was the coldest location within the building. The cold-est spot, Tsi, was measured as 19.3 �C. Te and Ti were 6.6 �C and 22.2 �C, respectively,which equates to fRsi = 0.81 and is above the critical temperature factor fCRsi foravoiding mould growth in dwellings (fCRsi = 0.75) and surface condensation in offi-ces or retail premises (fCRsi = 0.5) and sports halls, kitchens, or canteens (fCRsi =0.8).In addition to minimum temperatures at junctions, the temperatures throughthe depth of the wall and across the interior surfaces are an important parameterin determining the ability of the building fabric to moderate the diurnal and sea-sonal variations in the external environment and influence the thermal comfortof the occupants, which impacts energy use. Yates [16] observed that the temper-atures maintained in the Haverhill hemp houses were consistently one or 2� higherthan in the brick houses for the same amount of heat input. In the Hempod surfaceand intra-surface measurements of relative humidity and temperature were re-corded at the geometric centre of each wall. A timber moisture probe was fixed intoa timber strut in the centre of the wall and thermocouples attached to the internaland external wall surfaces. Hygrotrac� RHT sensors were embedded at 40 mmintervals within each wall. The installation of the embedded RHT and timber mois-ture sensors is presented in Fig. 3. Figs. 4 and 5 present the variation of internal andexternal air and surface temperatures, respectively, over a period of 11 days in mid-May 2011. Similarly, Fig. 6 presents internal and external relative humidity over thesame period. Fig. 7 presents internal air and surface temperatures. Figs. 4–6 indicate
3 The building was pre-heated prior to the data collection period.
Fig. 3. Installation of embedded RHT and timber moisture sensors within theHemPod wall.
Fig. 4. Internal and external air temperatures.
Fig. 5. Internal and external wall surface temperatures.
Fig. 6. Relative humidity of the internal and external air.
Fig. 7. Hempod internal air and wall surface temperatures.
A. Shea et al. / Construction and Building Materials 36 (2012) 270–275 273
that the temperature and humidity variations inside the Hempod are significantlydampened compared with the external environment. The ability of the buildingfabric to lessen the impact of external oscillations and to delay such effects is fun-damental to the avoidance of summertime overheating. A stable indoor environ-ment is clearly a desirable feature with regards to maintaining a satisfactory levelof occupant thermal comfort and is a characteristic more often associated withthermally heavyweight buildings. Analysis of temperature data for the entire monthof May reveals a mean daily swing in external temperature of 6.5 �C, whereas themean daily variation in the internal temperature is 0.9 �C equating to an attenua-tion of 86% (Fig. 8).
Fig. 8. Mean internal and external temperature swing.
Fig. 10. WUFI-simulated temperature profile from exterior (cold) side to interior(hot) side following a sudden temperature drop.
274 A. Shea et al. / Construction and Building Materials 36 (2012) 270–275
2.2. Laboratory experiment and WUFI simulation
The field test observations of Yates [16] and WUFI simulations presented in Evr-ard and De Herde [17] both reported that evaluation of hemp–lime building energyperformance through the use of the thermal transmittance (U-value) alone is notsufficient. Real buildings are subject to an external environment that is constantlychanging and, therefore, the building fabric rarely reaches a steady-state. Evrardand De Herde [17] identified that the dynamic behaviour of hemp–lime walls is bet-ter described by parameters that account for the transient behaviour of the mate-rial. Q24h and ts-s are two such parameters. Q24h (%) represents the energytransferred over 24 h as a proportion of that which would have been transferredif a steady-state had been reached instantaneously. The time to steady-state, ts-s
(h), is the time taken to reach 95% of the heat flow at the steady-state. A wall assem-bly with large ts-s is, therefore, not well represented by the steady-state U-value. Inorder to investigate these characteristics, a 300 mm thick test panel was con-structed and stored for approximately 12 months in an unconditioned internalspace. The panel was moved to the test laboratory and stored in a controlled envi-ronment (20�, 60% RH) for 30 days prior to the start of the tests. The designatedexterior of the test panel was enclosed within a temperature and humidity-con-trolled environmental chamber and the designated interior surface was exposedto temperature and humidity-controlled laboratory space. Relative humidity & tem-perature (RHT) measuring devices were cast into the panel at various depths. A sud-den drop in temperature was imposed by reducing the temperature in theenvironmental chamber by 20� to 0 �C, whilst the other side continued to be main-tained at 20 �C. Fig. 9 presents the temperature profile from exterior (cold) side tointerior (hot) side following the sudden temperature change. Many days elapsedbefore a steady-state began to be established. Taking a mean value from the closestto steady-state experimental data, the time to steady state ts-s is approximately240 h, which is considerably longer than expected. Q24h was 17% for a calculatedthermal transmittance of 0.3 W/m2 K. A wall of the same thermal resistance was de-fined in the transient heat and moisture simulation software WUFI Pro 5 and thesame thermal shock simulation conducted. Considering temperature only, ignoringthe effects of relative humidity, the simulated hemp–lime wall reaches a steady-state within 72 h (Fig. 10); considerably less time than the laboratory test that in-cludes the effects of phase-change within the material. Q24h for the simulated wallwas 19% similar to the laboratory measurements and simulations reported in Evr-ard and De Herde [17].
3. Discussion
The simple steady-state model designed to predict the whole-building heat loss for the Hempod test building produced a predic-tion close to the co-heating test results, which provides confidencein the use of such a test procedure for the evaluation of steady-state ‘as-built’ performance. However, it would be prudent to notethat there remains uncertainty in the true value of thermal conduc-tivity of the hemp–lime, due to differences in density and moisturecontent. Furthermore, the research team had the benefit of beingable to pre-heat the building for a two-week period prior to collect-ing data for the co-heating study. Additionally, estimation of theheat loss due to infiltration at normal pressures from blower doortests at higher pressures and varying environmental conditions
Fig. 9. Laboratory-measured temperature profile from exterior (cold) side tointerior (hot) side following a sudden temperature drop.
during field testing all contribute to error in the measurementand prediction the energy performance of the real building. Interms of relating these test results to the likely energy performanceof an occupied building, the co-heating test and steady-state modeldo not account for the effects of solar and internal heat gains and,consequently, any real building under normal operation will notrequire the heating system to meet the entire heat loss determinedby these methods. Yates [16] when comparing the performancehemp–lime construction with conventional brick-built homes atHaverhill, UK, found that the heating fuel consumed by the hemphomes was no greater than that used in the traditionally con-structed houses. However, SAP ratings and U-value calculationshad predicted that the hemp houses should have been using signif-icantly more energy than the brick houses.
Laboratory tests of the 300 mm un-rendered hemp–lime panelsubject to sudden cooling in a temperature and humidity-con-trolled environment show that the time taken to reach a steady-state was in the order of 10 days. The laboratory test was stoppedat 12 days. WUFI simulation of the same panel thickness and con-ductivity, ignoring the effects of relative humidity, predicts steady-state conditions in approximately 3 days, highlighting the influ-ence of relative humidity on the dynamic behaviour of hemp–lime.In both cases, it is evident that, as other authors have stated [17],the steady-state thermal transmittance alone is not able to repre-sent the transient behaviour of such materials.
4. Conclusions
Co-heating tests of a hemp–lime based test building haveshown good agreement with a simple steady-state heat loss pre-diction model. However, laboratory tests of a 300 mm hemp–limetest panel subject to a sudden 20 �C temperature drop took in theregion of 10 days to approach steady-state conditions. WUFI simu-lations of the same hemp–lime panel focussing on temperaturetook almost 3 days to reach a steady-state. In both simulationand laboratory tests, the time to reach a steady-state is many daysand this indicates that the U-value alone is, therefore, not a suit-able parameter for evaluating the thermal performance of hemp–lime walls subject to real, transient, weather conditions and shouldbe supplemented with evaluation of dynamic factors, for example,Q24h. The presented co-heating test results followed a two-weekperiod of pre-heating, which would serve to limit the effects ofthe thermal mass, although during the test diurnal variations inexternal air temperature were as high as 10 �C. Further work is re-quired to evaluate the influence of thermal mass and dynamichygrothermal behaviour on co-heating test data.
A. Shea et al. / Construction and Building Materials 36 (2012) 270–275 275
The results of temperature measurements show that the Hem-pod test building envelope provides a significant amount of atten-uation of the oscillations in the external environment which willassist in maintaining comfortable summertime conditions withinthe building. Blower-door tests demonstrate air permeability lev-els consistent with the demanding Passivhaus standard, which willsignificantly reduce energy use associated with winter-time infil-tration heat losses.
Acknowledgements
The Hempod research programme was funded by DEFRA undertheir LINK programme, and supported by the following industrialpartners: Lhoist (UK) Ltd., BRE, Hanson Ltd., Lime Technology Lim-ited, Hemp Technology Limited, Wates Construction Ltd., FeildenClegg Bradley studios, NNFCC. We also acknowledge the help andsupport of staff and students at BRE CICM, Department of Architec-ture and Civil Engineering at the University of Bath and the dona-tion of high performance door and windows by Janex Ltd.
References
[1] DECC, Department of energy and climate change statistical release, UK climatechange sustainable development indicator: 2010 greenhouse gas emissions,provisional figures and 2009 greenhouse gas emissions, final figures by fueltype and end-user; 31st March, 2011. <http://www.decc.gov.uk/assets/decc/Statistics/climate_change/1515-statrelease-ghg-emissions-31032011.pdf>[accessed 19.08.11].
[2] BIS, Innovation and growth team, low carbon construction – final report,department for business innovation and skills, London. <http://www.bis.gov.uk/assets/biscore/business-sectors/docs/l/10-1266-low-carbon-construction-igt-final-report.pdf> [accessed 19.09.11].
[3] Boutin MP, Flamin C, Quinton S, Gosse G. Analyse du cycle de vie de mur enbéton chanvre banché sur ossature en bois (Life cycle analysis of a cast hemp-lime timber framed wall). Lille: INRA; 2005.
[4] Cérézo, V. Propriétés mécaniques, thermiques et acoustiques d’un matériau àbase de particules végétales: approche expérimentale et modélisationthéorique. PhD Thesis, ENTPE, Lyon; 2005.
[5] Hills CAS, Norton AJ, Newman G. NFI – The importance of hygroscopicity inproviding indoor climate control. NOCMAT, Bath, UK; 6–9 September, 2009.
[6] Collet F. Caractérisation hydrique et thermique de matériaux de génie civil àfaibles impacts environnementaux. Thèse de Doctorat, INSA de Rennes; 2004.
[7] Evrard A.Transient hygrothermal behaviour of lime–hemp materials. PhD,Université Catholique De Louvain; 2008.
[8] Tran Le AD, Maaloufa C, Maia TH, Wurtzb E, Collet F. Transient hygrothermalbehaviour of a hemp concrete building envelope. Energy Build2010;42:1797–806.
[9] Papadopoulus AM. State of the art in thermal insulation materials and aims forfuture developments. Energy Build 2005;37:77–86.
[10] Dominguez-Munoz F, Anderson B, Cejudo-López JM, Carrillo-Andrés A.Uncertainty in the thermal conductivity of insulation materials. J EnergyBuild 2010;42:2159–68.
[11] BS EN ISO 6946:2007. Building components and building elements – thermalresistance and thermal transmittance – calculation method. London: BritishStandards Institution; 2007.
[12] Palmer J, Pane G, Bell M, Wingfield J. Comparing primary and secondary termsanalysis and re-normalisation (PStar) test and co-heating test results. A reportto the department for communities and local government. London: HMSO;2011 [ISBN: 978 1 4098 2902 7].
[13] ATTMA technical standard 1. Measuring air permeability of buildingenvelopes. Air tightness testing and measurement association, Issue 2, 13thJuly 2007. <http://www.attma.org> [accessed 06.11.11].
[14] BS EN 13829:2001. Thermal performance of buildings. Determination of airpermeability of buildings. Fan pressurization method. London: BritishStandards Institution; 2001.
[15] Ward TI. Assessing the effects of thermal bridging at junctions and aroundopenings. BRE IP 1/06, BRE, Watford; 2006.
[16] Yates T. Final report on the construction of the hemp houses at haverhill,suffolk, client report 209-717 Rev2, BRE, Watford; 2002.
[17] Evrard A, De Herde A. Hygrothermal performance of lime–hemp wallassemblies. J Build Phys 2010;34:5–25.