This article was downloaded by: [University of Western Ontario]On: 17 November 2014, At: 06:45Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Architectural Science ReviewPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/tasr20

Designing for thermal comfort near a glazed exteriorwallTimothy Anderson a & Mark Luther ba School of Engineering , Auckland University of Technology , Auckland , New Zealandb School of Architecture and Building , Deakin University , Geelong , AustraliaPublished online: 12 Jul 2012.

To cite this article: Timothy Anderson & Mark Luther (2012) Designing for thermal comfort near a glazed exterior wall,Architectural Science Review, 55:3, 186-195, DOI: 10.1080/00038628.2012.697863

To link to this article: http://dx.doi.org/10.1080/00038628.2012.697863

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Architectural Science ReviewVol. 55, No. 3, August 2012, 186–195

Designing for thermal comfort near a glazed exterior wall

Timothy Andersona∗ and Mark Lutherb

aSchool of Engineering, Auckland University of Technology, Auckland, New Zealand; bSchool of Architecture and Building,Deakin University, Geelong, Australia

In many highly glazed buildings, the thermal comfort of the occupants will tend to be related to the incoming solar energyand the heat transfer behaviour of the glazing. In this study, several glazing systems were designed using the software toolsVISION 3 (University of Waterloo 1992) and WINDOW-6 (Lawrence Berkeley National Laboratory 2011), with a view toimproving thermal environment of occupants near the glazed wall of a commercial office. The systems were fabricated andexperimentally tested to validate the software modelling results. Subsequently, the glazing systems were retro-fitted to theoffice and tested in situ for a summer month. Results of this testing, in the form of Fangers’ predicted mean vote (PMV) andthe predicted percentage dissatisfied (PPD), are presented, and some options for improving the thermal environment in thisnear-façade zone are discussed.

Keywords: design; glazing; modelling; thermal comfort; testing

IntroductionThe idea of human thermal comfort is one that is particularlycomplex; however, at its most basic it can be defined as ‘thatcondition of mind which expresses satisfaction with thethermal environment’ (ISO 7730 2005). Therefore, the ideaof someone being ‘comfortable’ relates not just to the airtemperature but also to a range of other non-environmentalparameters such as physiology, psychology and behaviouralfactors. As such comfort is a complex relationship betweenparameters such as metabolic rates, the level of clothingbeing worn, air temperature, relative humidity, mean radianttemperature, local air velocity and radiant asymmetry.

Fanger (1973) developed a quantitative method (the‘comfort equation’) to describe comfort levels using sixparameters as shown in

f (M , Clo, v, tr , ta, Pw) = 0 (1)

where M is the metabolic rate, Clo is a clothing index, v

is the air velocity (m/s), tr is the mean radiant tempera-ture (◦C), ta is the ambient air temperature (◦C) and Pw isthe vapour pressure of water in ambient air (Pa). Whenany combination of these parameters satisfies the com-fort equation the thermal comfort of the majority of theoccupants will be neutral, i.e. not too hot nor too cold.

In reality, it is not always possible to satisfy this criterionand so have ‘optimal’ thermal comfort. As such it is nec-essary to describe the relative level of discomfort. This canbe achieved by two quantitative means: the predicted meanvote (PMV) index and the predicted percentage dissatisfied

∗Corresponding author. Email: timothy.anderson@aut.ac.nz

(PPD) index (Fanger 1973). The PMV index predicts themean response of a large group of people according to theASHRAE thermal sensation scale and results in a ratingfrom +3 (hot) to −3 (cold). This PMV value is a functionof the metabolic rate (M ) and thermal load (Equation (2)),where the thermal load (L) is defined as the differencebetween the internal heat production and the heat loss to theactual environment for a person kept at comfortable valuesof skin temperature and evaporative heat loss by sweatingat their actual activity level:

PMV = (0.303e−0.036M + 0.028)L (2)

The second criterion, PPD, is a quantitative measure of thethermal comfort of a group of people in a particular thermalenvironment and is a more meaningful indicator of thermalcomfort. Fanger (1973) related the PPD to the PMV using

PPD = 100 − 95e−(0.03353 PMV2+0.2179 PMV2) (3)

What Fanger (1973) found from the PPD was that irre-spective of how perfect a thermal environment is, it is notpossible to achieve a PPD of less than 5% for similarlyclothed people undertaking the same activity.

Previously, work had been undertaken into the indoorenvironmental quality testing of a commercial office build-ing in a refurbished wool-store, shown in Figure 1. Duringthis investigation it was found that occupants near thesingle glazed eastern perimeter of the building reporteda higher degree of discomfort than those situated in the

ISSN 0003-8628 print/ISSN 1758-9622 online© 2012 Taylor & Francishttp://dx.doi.org/10.1080/00038628.2012.697863http://www.tandfonline.com

Dow

nloa

ded

by [

Uni

vers

ity o

f W

este

rn O

ntar

io]

at 0

6:45

17

Nov

embe

r 20

14

Architectural Science Review 187

Figure 1. Eastern glazed façade of office building.

Figure 2. Office floor plan showing area of interest.

interior workplace. One of the improvements made was theintroduction of an optimized conditioning control systemthat attempted to balance the temperature throughout theopen office space.

However, several years later, the management and occu-pants of this building remain dissatisfied with the thermalconditioning near the glazed perimeter, particularly onextremely hot days. As such, the building occupants wereinterested in investigating solutions to the problem thatcould improve thermal comfort, but not unduly disturb theaesthetic qualities of the existing façade.

Therefore, the objective of the study was to investi-gate possible solutions such that the occupants could workin a comfortable temperature range, particularly near theeastern perimeter (Figure 2) where there exists a 3-m floor-to-ceiling single-glazed façade with a selective mirror-facedexternal coating and a heat absorbing glass.

Given that previous attempts to rectify the situationin the building were aimed at modifying the air velocity(v), interior temperature (ta) and relative humidity (Pw) forFangers’ comfort equation, this study focused on address-ing the effect of the radiant temperature encountered by thebuildings occupants (tr in Fangers’ comfort equation).

Heat transfer in glazingGiven that occupants of the office near the glazed façade ofthe building complained of greater discomfort than those

eat transfer from absorbed solar

radiation by convection and

radiation

Heat transfer from absorbed solar radiation by convection and radiation

Transmitted solar radiation

Conduction

Heat transfer from absorbed solar radiation by convection and radiation

Figure 3. Heat transfer in a double-glazed window.

situated in the interior workplace, and the possible effectof the radiant temperature, it is important to understand themechanisms of heat transfer that are occurring to (and from)the glass and how they are defined.

The heat gains or losses to a glazing system are relatedto the physical properties of glass (conductivity, absorp-tion, reflection and transmittance) and are a function ofthe environmental conditions (solar radiation, wind andexternal/internal air temperatures). Predicting the heat gainor heat loss from a glass façade system is ultimately depen-dent upon the calculation of the glass surface temperature.In other words, after the physical properties are accountedfor with respect to the environmental conditions, we candetermine the glass’ surface temperature. Once the temper-ature is calculated, it is possible to calculate the amount ofheat transfer from, or to, the glazing by the buildings occu-pants by conduction, convection and radiation (Duffie andBeckman 2006).

The instantaneous transfer of heat to and from a glazingsystem is represented by three components as illustrated inFigure 3.

As such, the overall transfer of heat can then bedescribed by (ASHRAE 2005)

qA = Itτ + Ni(αIt) + U (to − ti) (4)

where qA is the total inward heat gain through the glass(W/m2), It is the incident solar radiation available on theexterior glass (W/m2), τ is the fraction of It which is directlytransmitted through the glass, Ni is the inward-flowing frac-tion of absorbed solar radiation, α is the fraction of It whichis absorbed within the glass, U is the total thermal con-ductance of the glazing unit (W/m2 K), to is the exterior oroutside (ambient) air temperature (K) and ti is the interiorair temperature (K).

In examining Equation (4), the coefficients of the firstand second terms (Itτ + Ni(αIt)) represent the solar heatgain coefficient (SHGC) of the glass. This is the transfer ofheat from the solar irradiated glass to the occupied space.The third term (U (to − ti)) represents the overall conductiveheat transfer through the glass considering the differences

Dow

nloa

ded

by [

Uni

vers

ity o

f W

este

rn O

ntar

io]

at 0

6:45

17

Nov

embe

r 20

14

188 T. Anderson and M. Luther

between the interior and exterior temperature only. In thisrespect, if the interior and exterior air temperatures areequal, the conductive term reduces to zero allowing onlythe heat gains due to radiation to occur.

Lastly, the ratio of the SHGC of a specific glazing unitto that of a standard double-strength A-grade (DSA) glass(a clear architectural glass 3mm thick) provides the shadingcoefficient. Usually, all other glazing systems are less thanunity when compared to a DSA glass and so low SCs arean indication of the glazing systems capability to reduceincident solar heat gain transfer into a building.

In summary, U-value, SHGC and shading coefficient(SC) could characterize the thermal comfort that may beencountered near glazing systems, as these coefficientsrelate to heat transfer through a glazing system.

Modelling heat transfer in the glass façadeIn light of the dissatisfaction with the thermal environmentnear the glass façade, and the possibility of it being linkedto the temperature and solar transmission of the glazingsystem, three alternatives were proposed to help remedythe issue.

Firstly, it was proposed that a shading screen fabricatedfrom a protruded 3-D micro-screen, situated on an externalframe, be installed 50mm from the existing glass façade.Secondly, it was suggested that an argon-filled insulatingglazing (IG) unit consisting of a clear and a low-e glassbe assembled on the interior of the existing façade, thusproviding a triple-glazed façade with the low-e surface onthe cavity side of the interior glass. Finally, the third solutionwas to utilize another argon-filled IG unit with the centreglass being a heat absorbing grey glass and again forminga tripled glazed low-e system.

The rationale for the three alternatives was to exploreoptions that could lead to a reduction of the mean radianttemperature from the glazing surface as well as the trans-mitted solar energy. In the first instance a shading devicewould block out large percentages of radiation reachingthe glass, yielding a lower surface temperature and result-ing in a cooler building air temperature, through reducedradiant and convective gains. As the shading device wasnot expected to lower the U-value of the glass it wasanticipated that the building might still be too cold dur-ing winter periods. However, shading offers to reduce solarheat gains in periods of extreme temperature (Tzempelikoset al. 2010).

In the second situation, triple-glazed units with a gas fillsuch as argon could provide an effective solution as suchunits would have a lower U-value. In addition, the introduc-tion of a low-e coating could reduce radiant heat transferfrom the window due to the higher infrared (long-wave)reflectance of the coated surface. Triple-glazed units havebeen found to minimize the overall inside air temperaturevariance of a building due to lower values of solar heat gainas well as loss (Fang et al. 2010). As such these units could

help improve thermal comfort under high radiation and alsoduring cold periods.

Having formulated three concepts for glazing toimprove thermal comfort it was decided to utilize two read-ily available glazing analysis programmes to determinethe effect of design modifications on the façades thermalperformance.

The first of these, VISION 3.0, was developed at theAdvanced Glazing System Laboratory at the University ofWaterloo in Canada (1992) and was used to predict theU-value, SHGC and SC for the proposed systems. In addi-tion, it was used to calculate the glass surface temperatures,absorbed radiation, transmitted radiation, the radiative andconvective component of heat flux as well as the transmittedvisible/solar ratio. The second software tool, WINDOW 6 apackage developed at Lawrence Berkeley National Labora-tory (2011) for analysing window thermal performance inaccordance with standard NRFC procedures, was also uti-lized in predicting the transient glass surface temperaturesin the façade system.

Original glazing systemThe original glazing system consisted of a single layer oftinted glass (TS30), the transmittance (τ ) and reflectance(ρ) of which are shown in Table 1, for the visible, solarand long-wave spectrums. It is interesting to note that theoriginal glazing system has a high emissivity value on itsinner surface. This would suggest that the incident radiationabsorbed by the glass would be radiatively transferred tooffice.

Figure 4 provides a graphical summary of the VISION3.0 predicted heat gains through the window (for the inter-nal and external temperatures and radiation shown). Fromthis, it can be seen that approximately 20% of the inci-dent radiation (147W/m2) is transmitted into the building,corresponding to the solar transmittance.

However, more importantly, it shows that the glassabsorbs over 60% of the incident solar radiation. This leadsto a significant increase in the glass temperature encourag-ing high rates of convective (C) heat transfer (66.3W/m2).Furthermore, the relatively high emissivity of the inner sur-face results in a large radiative (R) heat transfer rate into thebuilding (77W/m2).

What this modelling shows is that the original systemhas a high SHGC and SC. As such heat gains by the building

Table 1. Optical properties of single glazed system.

Long-waveVisible (%) Solar (%) reflectance (%)

Glazing system τ ρ τ ρ Out In

Original heatabsorbingtinted glass

29 17 21 16 16 35

Dow

nloa

ded

by [

Uni

vers

ity o

f W

este

rn O

ntar

io]

at 0

6:45

17

Nov

embe

r 20

14

Architectural Science Review 189

Figure 4. Heat transfer in single-glazed system.

Figure 5. Predicted transient temperature of originalsingle-glazed system.

will be high during high radiation times, and during colderperiods the building will suffer heat losses due to a highU-value and high emissivity. In turn if we consider thisover a range of conditions (varying radiation, temperatureand wind speed) using WINDOW 6, as shown in Figure 5,it can be seen that even at low radiation levels the glass’outside surface temperature may be significantly above theambient temperature.

Shaded glazing systemThe first proposed solution for modifying the glazed façadewas to provide a shading screen to the exterior of the glassfaçade. For the shaded glazing system the glass optical prop-erties are the same as for the original glazing system, as thescreen would be added to its exterior. However, for the sim-ulations, it was assumed that the screen would result in an85% decrease in the incident solar radiation reaching theglass.

Figure 6. Heat transfer in shaded single-glazed system.

Under this assumption, Figure 6 shows that by block-ing a significant portion of the radiation from reaching theglazing that the glass temperature is substantially reducedfrom that of the original glazing system. In turn, the totalheat gain is significantly less than the original glass and theinside glass surface temperature nearly 15◦C cooler thanthe ‘original’ glass system. The net result of the screenis that the shading screen should lead to improvements inthe SHGC and SC. However, the U-value of the systemis still relatively high, as the screen offers no real thermalresistance.

Clear triple-glazed systemThe second solution that was explored was a retrofit triple-glazing unit. In this system, the original glazing wasmaintained and a double-glazed unit, with a clear glass paneforming the middle pane of the triple-glazed unit, was fit-ted behind this on the interior of the building. As such, thisnew system had varying glass properties between the threelayers where the new middle and interior layers had highvalues of visible and solar transmittance and low values ofvisible and solar reflectance as shown in Table 2. The use ofa low-e coating on the outside of the interior layer of glassmeans the surface has a high long-wave reflectance whencompared to the other high emissivity surfaces.

Table 2. Optical properties of clear glass triple-glazed system.

Long-waveGlazing Visible (%) Solar (%) reflectance (%)system greytriple glazed τ ρ τ ρ Out In

Tinted glass 29 17 21 16 16 35Grey glass 32 10 29 8 16 16Low-e glass 82 10 66 22/10 84 16

Dow

nloa

ded

by [

Uni

vers

ity o

f W

este

rn O

ntar

io]

at 0

6:45

17

Nov

embe

r 20

14

190 T. Anderson and M. Luther

Figure 7. Heat transfer in clear triple-glazed system.

In Figure 7, with the clear triple-glazed system, it can beseen that a high amount of the incident radiation is absorbedin the outside layer of the glass. This means that the interiorsurfaces absorb very little radiation and therefore the insideglass temperature is much lower than the original system. Inthis system about 10% of the incident radiation is still trans-mitted, but because of the lower temperature of the internalpane, heat gain due to convection and radiation into theroom are significantly less. On this basis the overall systemshould perform better than the original system. Moreover,the clear triple glazing should have a much lower U-value,SHGC and SC than the original glazing system meaningthat it is reducing heat gain when ambient temperatures arehot and heat loss when ambient temperatures are cold.

Grey triple-glazed systemThe third solution that was explored was another retrofittriple-glazing unit. Again, in this system a double-glazedunit, with a grey glass pane forming the middle pane of thetriple-glazed unit, was fitted behind the original glazing onthe interior of the building. Table 3 shows the glass proper-ties for the three layers, where the grey middle layer has lowvisible and solar transmittance compared to the clear glassin the previous system. The interior glass however remainsa clear low-e glass, yielding high long-wave reflectance onthe outside of the interior glass.

Table 3. Optical properties of grey glass triple-glazed system.

Long-waveGlazing Visible (%) Solar (%) reflectance (%)system cleartriple glazed τ ρ τ ρ Out In

Tinted glass 29 17 21 16 16 35Clear glass 87 7 77 7 16 16Low-e glass 82 10 66 22/10 84 16

Figure 8 illustrates the grey triple-glazed system simu-lation results. In this system the middle grey glass layerabsorbs a significant fraction of the remaining radiationthrough the outside layer. This results in a low absorptionon the inside layer as a result of the low amount of transmit-ted radiation (<5%). However, the absorption by the greyglass leads it to having a higher inside temperature thanthe clear triple-glazed system and so a high rate of radiantheat transfer to the low-e pane. This leads to the inner panehaving a higher temperature thereby resulting in a higherconvective and radiant heat gain from the inside pane to theoffice space. In summary, this type of triple glazing shouldhave lower values of SHGC and SC than the clear IG unitdescribed previously but a higher U-value due to the radiantproperties of the grey layer of glass allowing more outwardheat loss.

Experimental setupIn order to further explore the findings from the model andits impact on thermal comfort, it was decided to conductan in situ experimental assessment of the three alternativeglazing systems in conjunction with the original system.

In collecting meteorological data, a weather stationassembled on the roof of an adjacent building was loggedfor the duration of the experiments. This station consistedof a pyranometer measuring global radiation on a horizon-tal plane and a HMP35C Temperature/Relative HumidityProbe to measure the air temperature, relative humidity andwind speed. These data were logged at 15min intervals usinga CR10X Campbell Scientific data logger.

Additionally, a LI-COR silicon pyranometer, wasmounted on the façade of the building to measure the solarradiation incident on the windows. An array of calibrated(±0.3K) light gauge T-type thermocouples prepared for

Figure 8. Heat transfer in grey glass triple-glazed system.

Dow

nloa

ded

by [

Uni

vers

ity o

f W

este

rn O

ntar

io]

at 0

6:45

17

Nov

embe

r 20

14

Architectural Science Review 191

solar exposure, as suggested by Liu (1988), was used tomeasure the surface temperature of the glazing systems.Further radiant asymmetry sensors were placed approx-imately 500mm from the inside glazing to measure theradiant heat gain or loss from the windows.

To assess the comfort level of the occupants a 1221Bruel & Kjaer Comfort Meter, placed 500mm from theglazing, was used to measure the interior glass surfacetemperature, dry and wet bulb temperature, mean radianttemperature, absolute humidity, air velocity, operative tem-perature and the radiant asymmetry. The outputs from thesewere used in determining thermal comfort parameters setout in ISO-7730 (Fanger Model of comfort).

Modelling validationHaving undertaken design calculations of the façade ther-mal performance using steady-state software modellingtools, it was necessary to validate the findings from thesemodels, and test them under transient conditions. As suchthe temperatures predicted by Vision and Window wereexamined, and found to be comparable. Subsequently, theexperimentally measured incident solar radiation, ambienttemperature, wind speed and internal air temperature fora range of conditions were input into WINDOW 6. FromFigure 9 it can be seen that there was good correlationbetween the modelled and measured temperature on theinside of the original glazing system.

Similarly, Figure 10 shows good correlation betweenthe modelled and measured inside glass surface temperatureof the clear triple-glazed system, for most conditions. Thissuggests that even for multilayered glazing systems, steady-state models are able to achieve a relatively satisfactorycorrelation with reality.

On this basis it can be concluded that the findings ofthe modelling are likely to be borne out in practice. Thosebeing that the radiant heat transfer from the original glazingwill be a large fraction of the inward flowing heat gain andthat the use of low-e glasses can reduce the radiant transferfrom the glass to the interior.

Figure 9. Predicted versus measured glass temperatures insidesurface of original glazing system.

Figure 10. Predicted versus measured glass temperatures insidesurface of clear triple glazing.

ResultsTo quantify the indoor comfort, a PMV and PPD calcula-tion was made for each of the glazing options. This could beutilized to determine when the occupants were the least sat-isfied, and if the occupants were dissatisfied, was it becausethey perceived the building to be too cold or too hot. Asthe PPD is calculated from the measured values from thecomfort meter including dry bulb temperature and air veloc-ity, the instrument was placed adjacent to the east facingfaçade.

Original glazing systemBefore exploring the potential solutions, a benchmarkassessment of the PPD was performed on the existing glazedfaçade assuming a clothing index of 0.8 and a metabolic rateof 1.0, typical values for Australian office workers (Rowe2001), though perhaps conservative for summer conditions.That said, it should be noted that if the occupants clothingindex were lower (0.65) given summer periods, and they hada higher metabolic rate (1.2) the results are comparable.

Figure 11 illustrates a period of warm weather withhigh incident radiation where the office air temperature nearthe façade reached nearly 30◦C. Such conditions cause theoccupants to be quite dissatisfied, with around 30% pre-dicted to be dissatisfied with the thermal comfort in thebuilding. In addition it can be seen that relatively smallchanges in the inside temperature often lead to dramaticchanges to the PPD. This illustrates that the original glaz-ing system is not aiding the thermal environment, and thatduring hot summer conditions a large percentage of theemployees would be dissatisfied.

More importantly, before midday tends to be the worsttime of day, as this is when the eastern face of the buildingis under high levels of incident solar radiation. However,once the building has been heated in the morning it stayswarm all day, leaving occupants dissatisfied all day long.Furthermore, if the weather cools down it can be seen thatthe office can be perceived to be too cold.

Dow

nloa

ded

by [

Uni

vers

ity o

f W

este

rn O

ntar

io]

at 0

6:45

17

Nov

embe

r 20

14

192 T. Anderson and M. Luther

0

100

200

300

400

500

600

700

800

0

5

10

15

20

25

30

35

40

Inci

dent

Sol

ar R

adia

tion

(W/m

2 )

Am

bien

t Tem

pera

ture

(°C

)/P

redi

cted

Per

cent

age

Dis

satis

fied

(%)

Time

Indoor TemperatureAmbient TemperaturePPDIncident Radiation

TooHot

TooCold

12:0012:0012:0012:0012:00

Figure 11. Predicted percentage dissatisfied (original glazing).

–3

–2

–1

0

1

2

3

6:00

6:30

7:00

7:30

8:00

8:30

9:00

9:30

10:00

10:30

11:00

11:30

12:00

12:30

Pre

dict

ed M

ean

Vot

e

Time

80% SatisfiedSeries5Vote

hot

warm

slightly cool

cool

cold

slightly warm

neutral

0.8CLO1.0 MET

Figure 12. Predicted mean vote (shaded glazing).

Shaded glazing systemAs mentioned previously, a PMV and PPD calculation wasmade for each alternative system explored during this study.When examining, the PMV for the fixed shaded glazingsystem (Figure 12), it can be seen that the occupants arelikely to be satisfied with their environment. It is interest-ing to note that during early office hours (c. 8:30) there is aperiod where occupants may feel slightly warm. This is alsoillustrated in the PPD assessment, Figure 13, where a briefperiod of increased dissatisfaction is predicted. However,it can be seen that as the indoor air temperature decreases,the PMV and PPD also decrease, suggesting the thermalcomfort for the occupants with this glazing arrangementwill be dominated by the mean radiant and indoor air tem-perature. Similarly, it is important to note that during thevery early morning, the shaded system experiences a nega-tive radiant asymmetry (radiant heat loss from the building).

This suggests that during cooler periods, the shaded glazingsystem could lead building occupants near the façade to feelcold.

Clear triple-glazed systemWhen examining, the PMV for the clear triple-glazing sys-tem (Figure 14), it can be seen that the occupants are likely tofeel slightly warm in their environment. Again in Figure 15it can be seen that as the indoor air temperature decreases,the PMV and PPD also decrease, however it is interest-ing to observe that at approximately 9:00, there is a sharpincrease in the predicted dissatisfaction, with only a slightchange in the radiant asymmetry and indoor air tempera-ture. This suggests that for this instance, the air temperaturebecomes less significant than other factors, such as radiantheat transfer from the glass. On the whole, however, the

Dow

nloa

ded

by [

Uni

vers

ity o

f W

este

rn O

ntar

io]

at 0

6:45

17

Nov

embe

r 20

14

Architectural Science Review 193

–100

0

100

200

300

400

500

600

700

800

900

0

5

10

15

20

25

30

35

40

45

50

Inci

dent

Sol

ar R

adia

tion/

Rad

iant

Asy

mm

etry

(W

/m2 )

Am

bien

t Te

mpe

ratu

re (

°C)/

Pre

dict

ed P

erce

ntag

e D

issa

tisfie

d (%

)

Time

Mean Radiant Temperature

Indoor Temperature

Ambient Temperature

PPD

Incident Radiation

Radiant Asymmetry

13:1212:0010:489:368:247:126:004:48

Figure 13. Predicted percentage dissatisfied (shaded glazing).

–3

–2

–1

0

1

2

3

6:00

6:30

7:00

7:30

8:00

8:30

9:00

9:30

10:00

10:30

11:00

11:30

12:00

12:30

13:00P

redi

cted

Mea

n V

ote

Time

80% Satisfied

Vote

hot

warm

slightly cool

cool

cold

slightly warm

neutral

0.8CLO1.0 MET

Figure 14. Predicted mean vote (clear triple glazing).

0

100

200

300

400

500

600

700

800

900

1000

0

10

20

30

40

50

60

70

80

90

100

Inci

dent

Sol

ar R

adia

tion/

Rad

iant

Asy

mm

etry

(W

/m2 )

Am

bien

t Tem

pera

ture

(°C

)/P

redi

cted

Per

cent

age

Dis

satis

fied

(%)

Time

PPD

Mean Radiant Temperature

Indoor Temperature

Ambient Temperature

Incident Radiation

Radiant Asymmetry

14:2413:1212:0010:489:368:247:126:004:48

Figure 15. Predicted percentage dissatisfied (clear triple glazing).

Dow

nloa

ded

by [

Uni

vers

ity o

f W

este

rn O

ntar

io]

at 0

6:45

17

Nov

embe

r 20

14

194 T. Anderson and M. Luther

–3

–2

–1

0

1

2

3

6:00

6:30

7:00

7:30

8:00

8:30

9:00

9:30

10:00

10:30

11:00

11:30

12:00

12:30

13:00P

redi

cted

Mea

n V

ote

Time

80% Satisfied

Vote

hot

warm

slightly cool

cool

cold

slightly warm

neutral

0.8CLO1.0 MET

Figure 16. Predicted mean vote (grey triple glazing).

-100

0

100

200

300

400

500

600

700

800

900

0

5

10

15

20

25

30

35

40

45

50

Inci

dent

Sol

ar R

adia

tion/

Rad

iant

Asy

mm

etry

(W

/m2 )

Am

bien

t Tem

pera

ture

(°C

)/P

redi

cted

Per

cent

age

Dis

satis

fied

(%)

Time

Ambient TemperatureMean Radiant TemperaturePPDIndoor TemperatureIncident RadiationRadiant Asymmetry

14:2413:1210:48 12:009:368:247:126:004:48

Figure 17. Predicted percentage dissatisfied (grey triple glazing).

relatively high temperature in the office suggests that airtemperature is still a significant factor, and that perhaps thecooling system was not functioning appropriately.

Grey triple-glazed systemWhen examining, the PMV for the grey triple-glazing sys-tem (Figure 16), it can be seen that the occupants are againlikely to feel slightly warm in their environment. In Fig-ures 16 and 17 it can be seen that as the mean radiant andindoor air temperature increase, the PMV and PPD alsoincrease, as would be expected.

However beyond 25◦C the air temperature becomes lesssignificant than other factors. This is most clearly observedin Figure 17, when at approximately 9:00, the indoor airtemperature is constant, but the radiation level is still rela-tively high and the outside temperature is increasing. Suchconditions lead to a reduction of heat loss to the surround-ings, which in turn leads to an increase of heat gain byoccupants for a short period of time. However, it is perhaps

more important to note that the largest predicted increasein dissatisfaction occurs for the period where the meanradiant and indoor temperature increases above 25◦C to itsmaximum (7:12 to 8:24).

DiscussionWhen examining the alternative solutions to improvingthermal comfort in the office building, there are a num-ber of clear findings. Firstly, the indoor air temperatureappears to be the determining parameter of thermal com-fort for all glazing systems up to 25◦C. Beyond this level,the incident radiation and outside temperature show someeffect on thermal comfort but this appears to be less sig-nificant. For example, in the prediction of dissatisfactionbehind the grey triple-glazed façade, it was seen that withthe indoor air temperature increasing from 25◦C to approxi-mately 27◦C, the PPD increased by over 20%. However, forthe same glazing with the indoor air temperature constant at

Dow

nloa

ded

by [

Uni

vers

ity o

f W

este

rn O

ntar

io]

at 0

6:45

17

Nov

embe

r 20

14

Architectural Science Review 195

approximately 27◦C, a reduction in incident radiation from450 to 100W/m2 improved satisfaction by 10%.

Similarly, for the shaded system, with the indoor airtemperature being almost constant at 25◦C, there was a highdegree of satisfaction predicted, whereas, in the case of cleartriple glazing, the high indoor temperature would suggesta high degree of thermal discomfort. As such control ofthe indoor temperature is paramount; however, reducingthe incoming radiation, a driving factor for increased airtemperatures, helps reduce the load that must be met bybuilding services. In this regard the three tested systemshelp achieve this result.

Considering both the modelled and measured results, itis possible to draw some conclusions about the glazing sys-tems. Firstly, a system such as the shading screen could leadto improvements in thermal comfort (and energy efficiency)during summer periods, by reducing the solar gain throughthe windows. However, during cooler periods it would beineffective in preventing heat losses from the building andcould lead to the occupants feeling cold.

Conversely, the use of triple-glazed systems may beless effective in controlling solar and radiant heat gains;however, they offer better insulating characteristics. Whencoupled with cooling in summer and heating in winter itis probable that the triple-glazed systems would lead togreater stability in the thermal environment. Therefore,by reducing the fluctuations in the thermal environment itmay be possible to achieve increased satisfaction from theoccupants.

ConclusionFor many highly glazed buildings in Australia the thermalcomfort of the occupants will be related to the incomingsolar energy incorporated in the SHGC of the glazing. Asconventional buildings tend to exhibit deep floor plans rela-tive to their height, areas close to the façade receive a muchgreater amount of the incoming energy. This imbalanceleads to occupants near the façade experiencing high dissat-isfaction with their thermal environment in the ‘near-façadezone’.

This study utilized two software tools, VISION 3 (Uni-versity of Waterloo 1992) and WINDOW-6 (LawrenceBerkeley National Laboratory 2011), to analyse and designa retrofit glazing system with a view to improving ther-mal environment of occupants near the glazed wall ofa commercial office. In situ testing of the glazing sys-tems demonstrated that the simulation models provide arelatively good prediction of glazing temperatures for com-mon glazing arrangements, and as such the heat transfercomponents.

Furthermore, the façade testing instrumentation alloweda quantitative analysis of the thermal environment using

Fangers’ PMV and the PPD. The findings of the modelling,testing and thermal comfort calculation provided someresults leading towards improving the thermal environmentin this ‘near-façade zone’. It was found that control ofthe indoor temperature was paramount; however, it alsoshowed that there are advantages to being able to reducethe incoming radiation directly, as well as to reduce theinterior glass surface temperature through the use of staticshading systems and advanced glazing systems.

EpilogueSince this study was undertaken, a retrofitting of the greytriple-glazing system to the east façade has occurred, asrequested by the offices’ occupants. Anecdotal reports indi-cate that occupants are now comfortable with typical hotsummer conditions and that the interior shading is onlyused to prevent glare on the computer screens for a fewearly hours. Furthermore, the ‘near façade zone’ that waspreviously used as a walkway is now fully occupied byworkstations butting the glass façade. The buildings occu-pants have also reported that the new glazing installationhas reduced the burden on the HVAC system and thatdesired set-point conditioning takes approximately 15minto achieve.

ReferencesASHRAE, 2005. 2005 ASHRAE handbook – fundamentals.

Atlanta: American Society of Heating, Refrigerating, andAir-Conditioning Engineers Inc.

Duffie, J.A. and Beckman, W.A., 2006. Solar engineering ofthermal processes. New York: Wiley.

Fang, Y., Hyde, T.J., and Hewitt, N., 2010. Predicted thermalperformance of glazing. Newtownabbey: School of the BuiltEnvironment, University of Ulster.

Fanger, P.O., 1973. Thermal comfort. New York: McGraw-Hill.International Standards Organisation (ISO), 2005. ISO 7730:2005

Ergonomics of the thermal environment – analytical deter-mination and interpretation of thermal comfort using cal-culation of the PMV and PPD indices and local thermalcomfort criteria. Geneva: International Organization forStandardization.

Lawrence Berkeley National Laboratory, 2011 [online]. Availablefrom: http://windows.lbl.gov/software/window/6/index.html[Accessed 11 August 2011].

Liu, S.T., 1988. Temperature measurement of glass subjected tosolar radiation. ASHRAE Transactions, 94 (2), 1359–1369.

Rowe, D.M., 2001. Activity rates and thermal comfort of officeoccupants in Sydney. Journal of Thermal Biology, 26 (4–5),415–418.

Tzempelikos, A., et al., 2010. Indoor thermal environmental con-ditions near glazed facades with shading devices – Part II:thermal comfort simulation and impact of glazing and shad-ing properties. Energy and Building, 45 (11), 2517–2525.

University of Waterloo, 1992. VISION 3.0, glazing system thermalanalysis. Waterloo: Advanced Glazing System Laboratory,University of Waterloo.

Dow

nloa

ded

by [

Uni

vers

ity o

f W

este

rn O

ntar

io]

at 0

6:45

17

Nov

embe

r 20

14