8/10/2019 126712

1/8

2.7 URBAN DESIGN AND THERMAL COMFORT:ASSESSMENT OF OPEN SPACES IN BARRA FUNDA, A BROWNFIELD SITE IN

SO PAULO, BY MEANS OF SITE MEASUREMENTS AND PREDICTIVE SIMULATIONS

Denise Duarte, Joana Goncalves, Leonardo M. MonteiroUniversity of Sao Paulo, Sao Paulo, Brazil

1. INTRODUCTION

Continuing previous works, thisresearch focus on thermal comfort conditionsin different microclimates influenced by theurban design of Barra Funda, a neighborhoodin the west side of the city of Sao Paulo, Brazil.Recently, this area was object of investigation

for British and Brazilian specialists, includingthe authors, aiming to explore possibilities ofintervention for the proposal of urban realm.Duarte and Goncalves (2006) presented thestudy field, contextualizing it in the city of SaoPaulo, and discussing questions related tovegetation, morphology, building materials anddensity. In this research, the objective is toverify the adequacy of different configurationsof urban outdoor spaces in Barra Funda,inferring possible correlations between thermal

sensations and morphological characteristics,including vegetation, of each outdoor spacestudied.

2. EMPIRICAL RESEARCH

The adopted method is empirical, bymeans of field researches consideringmicroclimatic variables, and analytical,performing computational simulations ofpredictive models. The field research was doneduring a summer day, at seven different

locations, in a walking distance from eachother, as seen in Figure 1. The spots indicatedin Figure 1 can be seen in detail in Figure 2.Figure 3 brings the sky view factor and thenebulosity conditions for each one of the pointsof study. In each base, there was a set with athermohigrometer Homis model 229 and ananemometer Homis model 209. The data wasregistered in fifteen minutes intervals, between7am/8am, 10am/11am, 1pm/2pm and4pm/5pm. The data from the meteorological

station (Public Health Faculty of University ofSao Paulo), located 2,7 km far from the area,was also considered in the data assessments.

Following, one may find the procedures to themeasurement and treatment of the consideredvariables.

Figure 1: Site of study and points ofmeasurements.

2.1 Air temperature and humidity

Air temperature is a physical quantity that canbe measured by several methods, dependingon the kind of the sensor that is being used,which temperature can differ from the airtemperature due to radiant effects. Therefore,

the sensor should be protected from thermalradiation without compromising the properventilation in the surroundings of the sensor.ISO 7726 (1998) mentions three strategies toreduce the radiation effect. The first one isreducing the emission factor of the sensor,polishing it or using reflective painting; thesecond is reducing the temperature differencebetween the sensor and the surroundingsusing reflective surfaces of 0,1/0,2mm ofthickness; the last one is increasing the

convection coefficient, increasing mechanicallythe air speed around the sensor and/orreducing the sensor size.

1

SpotsTransects

gua Branca ParkLatin Am. MemorialBarra Funda Station

8/10/2019 126712

2/8

Figure 2: Spots of measurements andtransects.

In this research, digital thermohigrometersHomis 229 were used, which were calibratedcomparing their results with a calibrated one.The sensors were protected from direct solarradiation by means of polystyrene with externalaluminized surface and perforations of betterconvection thermal exchanges. Themeasurements were taken at 1,1m of height.The sensors used for quantifying the airtemperature are semiconductors, with readingrange of -20 C / +60 C, resolution of 0,1 C,precision of 0,4 C and response time of 0,1C/s. The humidity sensors are of capacitance,obtaining the relative humidity. The readingrange from 10% to 95%, with resolution of0,1%, precision of 3% (at 25 C, between30% and 95%) and 5% (at 25 C, between

10% and 30%) and response time of 3 minutes(from 45% to 95%) and 5 minutes (from 95% to45%). Considering the relative humidity (rh)and the air temperature (ta), the partial vaporpressure (pv) were obtained.

pv= 6,112 10^[7,5ta/ (237,7 + ta)]0,01 rh (01)

2.2 Wind speed

Air velocity is a physical quantitydescribed by its magnitude and direction. ISO7726 (1998) indicates the need of consideringthe direction of the flow and the magnitudefluctuation, and also the mean value and thestandard deviation during a certain period oftime. The helix sensors used present readingrange of 0,4m/s-30m/s, resolution of 0,1m/s,and precision of 2%+d, in which d is theobserved value. The data was registered in aten second interval during one minute, everyfifteen minutes. As the helix sensors areunidirectional, the wind direction wasdetermined by a light weight stream. The windspeed (va) considered is the arithmetical meanvalue of the instantaneous values obtained. Inorder to consider the magnitude fluctuation, thestandard deviation (sd) is estimated, where: n= number of instantaneous measurements,considering from i = 1 till i = n. Theturbulence intensity (it) is given in percentage.

dp = [1/(n-1)]1/2 n i=1(va,i- va)2 (02)

it = 100 sd/va (03)

P4

P1

P4

P3-P4

P3

P2

P5

P6

P7

P1-P2

8/10/2019 126712

3/8

Frequncia de ocorrncia de ventos

0

10

20

30

40

50N

NE

E

SE

S

SO

O

NO

2.3 Solar radiation

Figure 3 brings the sky view factor andthe nebulosity conditions for the points ofstudy.

Figure 3: Sky view factor e nebulosityconditions.

In order to consider the solar radiation(Rc) that will be absorbed by the body,Blazejczyk (2002) model was used, whichestimates the solar radiation considering thesolar height (h) and nebulosity (N), which isbetween 0 and 1. The variable a' is function of

the albedo (a) of the skin/clothing ensemble.An albedo value of 0,5 was adopted.

h 4:Rc = 1,4 a' (1,388 + 0,215 h)^2

h > 4 and N 0,20:Rc = 1.4 a' (-100,428 + 73.981 ln(h))

h > 4 and N = 0,21-0,50:Rc = 1,4 a' exp(5.383 16,072 / h)

h > 4 and N = 0,51-0,80:Rc = 1,4 a' exp(5.012 11,805 / h)

h > 4 and N > 0,80:Rc = a' 0,9506 h^1,039

(04)

(05)

(06)

(07)

(08)

a' = 1 0,01 a (09)

2.4 Empirical research results

Figure 4 shows the empirical datagathered and treated. These results are

considered to perform the predictivesimulations.

P1 frum

P2 vila/frum

P3 memorial

P4 matarazzo/memorial

P5 matarazzo

P6 arena

P7 parque

Estao (FSP)

Figure 4: Microclimatic data

P1: Low rising buildings streetP2: Frum SquareP3: Latin America Latina MemorialP4: Francisco Matarazzo AvenueP5: Agua Funda Park GateP6: Agua Funda Park ArenaP7: Agua Funda Park

Trees

Meteorological Station (FSP)

Umidade Relativa

20

30

40

50

60

70

80

90

100

07:00

07:15

07:30

07:45

08:00

10:00

10:15

10:30

10:45

11:00

13:00

13:15

13:30

13:45

14:00

16:00

16:15

16:30

16:45

17:00

%Temperatura do ar

18,0

20,0

22,0

24,0

26,0

28,0

30,0

32,0

34,0

07:00

07:15

07:30

07:45

08:00

10:00

10:15

10:30

10:45

11:00

13:00

13:15

13:30

13:45

14:00

16:00

16:15

16:30

16:45

17:00

Cair temperature

Radiao Solar

0

100

200

300

400

500

600

700

800

900

1000

07:00

07:15

07:30

07:45

08:00

10:00

10:15

10:30

10:45

11:00

13:00

13:15

13:30

13:45

14:00

16:00

16:15

16:30

16:45

17:00

W/m solar radiation

Velocidade do vento

0,00

0,50

1,00

1,50

2,00

2,50

3,00

3,50

4,00

07:00

07:15

07:30

07:45

08:00

10:00

10:15

10:30

10:45

11:00

13:00

13:15

13:30

13:45

14:00

16:00

16:15

16:30

16:45

17:00

m/s wind speedVelocidade do vento

0,00

0,50

1,00

1,50

2,00

2,50

3,00

3,50

4,00

07:00

07:15

07:30

07:45

08:00

10:00

10:15

10:30

10:45

11:00

13:00

13:15

13:30

13:45

14:00

16:00

16:15

16:30

16:45

17:00

m/s wind speed

wind direction

8/10/2019 126712

4/8

3. SIMULATIONS

In order to perform the computationalsimulations, the thermo physiological balancewas adopted, which can be represented asfollowing:

S = M + L + Rc + C + K + E sk+ Res (10)

Where: S: heat storage in the body, M: metabolic

power, Q: heat exchange on the skin by radiation,

C: heat exchange on the skin by convection, K: heat

exchange on the skin by conduction, Esk: heat loss

by evaporation at skin surface, Res: respiration heat

loss, all of them in [W/m]

In this research a metabolic rate

M=130W/m2 was adopted, equivalent to

walking at 1 m/s, or 3,6km/h, according ISO8996 (1990). Considering clothing, a thermalinsulation Icl=0,5, equivalent to light trousers,short sleeves shirt, underware and light shoes,was considered, following ISO 9920 (1995).

In the next topics, the followingestimations will be considered: (1) meanradiant temperature (trm), (2) superficial skintemperature, and (3) heat load index, in orderto assess the thermal conditions of the outdoor

spaces in study.

3.1 Mean radiant temperature

The mean radiant temperature (trm) isthe uniform surface temperature of a blackenclosure with which an individual exchangesthe same heat by radiation as the actualenvironment considered (ASHRAE, 2001). Itdescribes the radiant environment for a point inspace. It can be estimated by means of theglobe temperature, the two-spheres

radiometer, the constant temperature sensor orby the superficial temperatures and anglefactors.

In this research, the mean radianttemperature was calculated considering thethermal radiative exchanges between thehuman body and the sky (Lc) and thesurroundings (Lp), considering, as well, theexposure to solar radiation (Rc). In order todetermine the surrounding temperature (Tp),air-sun temperature was used, which can bedefined as the external superficial temperaturethat gives the same thermal effect than the

combination of the air temperature and theincident thermal radiation.

trm = [(Rc + 0,5 Lc + 0,5 Lp) / (0,95 5,66710^-8)]^0,25 - 273 (10)

Lc = 5,5 10^-8 (273 + t)^4 [0,82 0,25 10^(-0,094 0,75 vp)]

Lp = 5,5 10^-8 (273 + Tp)^4

(11)

(12)

3.2 Skin superf icial temperature

In order to estimate the superficial skintemperature, the following equation proposedby Blazejczyk (2004) was applied.

Ts= (26,4 + 0,02138 Trm + 0,2095 t 0,0185 ur 0,009 v) + 0,6 (Icl - 1) + 0,00128M (13)

3.3 Heat Load Index

Blazejczyk (2002) proposes theMENEX, Man-ENvironment heat EXchangemodel, wich uses the body thermal balance. Itsspecificities are: evaporative loss pondered bysex (1.0 for men; 0.8 for women), radiationexchanges pondered by nebulosity, solarradiation possibly considered by three differentmodels: SolDir, which considersdirect, diffuse

and reflect solar radiation; SolGlob, whichconsiders global solar radiation; SolAlt, whichcan be used when there is no solar radiationdata. These models consider clothing thermalresistance and pondered albedo of skin andclothes, presenting different equationsaccording to solar elevation and nebulosity. Inthis research, the third model was used; theother two models can be found in Blazejczyk(2004). In order to evaluate the results, it isproposed the criterion based on the Heat Load(HL), which is assessed considering the heat

storage (S), absorbed solar radiation (RC) andevaporative skin losses (Esk).

S 0 W/m2and Esk-50 W/m2

HL = [(S + 360) / 360] [ 2 - 1/(1+Rc)]

S > 0 W/m2and Esk-50 W/m2

HL = [(S + 360) / 360] [ 2 + 1/(1+Rc)]

S 0 W/m2and Esk< -50 W/m2

HL = (Esk/-50) [(S + 360) / 360][ 2 - 1/(1+Rc)]

S > 0 W/m2and Esk< -50 W/m2

HL = (Esk/-50) [(S + 360) / 360][ 2 +1/(1+Rc)]

(14)

(15)

(16)

(17)

8/10/2019 126712

5/8

3.4 Simulation results

Figure 5 presents the simulationresults, bringing the values of mean radianttemperature, of skin superficial temperatureand of heat load index, for each one of the

study points and for the data from themeteorological station.

P1 frum

P2 vila/frum

P3 memorial

P4 matarazzo/memorial

P5 matarazzo

P6 arena

P7 parque

Estao (FSP)

Figure 4: Estimated data

4. EMPIRICAL VERIFICATION/CALIBRATION

Monteiro and Alucci (2006) presentempirical research that verifies differentoutdoor thermal comfort model and indexes.One of the indexes that presented better

correlations with the empirical data gatheredwas Heat Load (correlation of 0,88 for themodel parameter and 0,83 for theinterpretation of the index). It is observed thatthe index values presents better correlationthat the others, although the interpretationcorrelation is close to them. On the other hand,as the correlation is better, it is possible toachieve more significant results, by means ofthe proposal of new interpretative ranges forthe index. The same authors (2007) present

results of a research in which newinterpretative ranges are proposed based onempirical researches carried on the city of SaoPaulo, considering over 70 microclimaticsituations and 2000 applied questionnaires.Due to the existence of such calibration donespecifically to a population adapted to SaoPaulo climatic conditions, the Heat Load Indexwas chosen as the criterion for interpreting theresults of this research. Table 1 brings theoriginal index (Blazejczyk, 2002) and theproposed calibration by Monteiro and Alucci(2007). The last one was adopted here, as theempirical base was gathered in So Paulo.

Table 1: HL (Blazejczyk, 2002) and calibrati onproposed by Monteiro & Alucc i (2007).

HL Classific ation Calibration Sensation

1,65 very hot

1,600 high stress (hot) 1,23-1,65 hot

1,186-1,600 stress (hot) 1,08-1,23 warm

0,931-1,185 neutrality 0,88-1,08 neutral

0,811-0,930 stress (cold) 0,72-0,88 cool0,810 high stress (cold) 0,65-0,72 cold

0,65 very cold

Other indexes, also calibrated byMonteiro and Alucci (2007), which alsopresented good results, should be used here,as well. Nevertheless, the Heat Load indexwas chosen since all the necessary input datawas available or was able to be estimatedusing the MENEX model.

P1: Low rising buildings streetP2: Frum SquareP3: Latin America Latina MemorialP4: Francisco Matarazzo AvenueP5: Agua Funda Park GateP6: Agua Funda Park ArenaP7: Agua Funda Park

Trees

Meteorological Station (FSP)

8/10/2019 126712

6/8

5. FINAL RESULTS

Figure 6 presents the final results of thisresearch, considering the thermal sensationsfound in each spot of study. In this Figure, Nstands for neutrality, W for warm, H for hot,

and VH for very hot.

HL P1 P2 P3 P4 P5 P6 P7 MS

07:00 N N N N N N N N

07:15 N N N N N N N N

07:30 N N N N N N N N

07:45 N N N N N N N N

08:00 N N N N N N N N

10:00 W W H W W W N H

10:15 H H H H W H N H10:30 H H H H W H N H

10:45 H H H H W H N H

11:00 H H H H H H N H

13:00 VH VH VH VH VH VH H VH

13:15 VH VH VH VH VH VH H VH

13:30 VH VH VH VH VH VH H H

13:45 VH VH VH VH VH VH H H

14:00 VH VH VH VH VH VH H H

16:00 VH VH VH VH VH VH H W16:15 VH VH VH VH VH VH H N

16:30 VH H VH VH VH VH H N

16:45 H H VH VH VH VH H N

17:00 H H VH VH VH VH H N

Figure 6: Final results of thermal sensation forthe seven points of study and for t he

meteorological station.

One may observe in Figure 6,

considering the first period of measurements,from 7am to 8am, that all points of studypresent similar thermal performance, withneutral thermal sensations observed (In Figure5, one may also notice that all the studiedpoints presented heat load values close to theunity).

During the second period ofmeasurements, from 10 am to 11 am, one mayobserve the change from the predominant

sensation of warm to hot. Two exceptionsmust be considered, both of them at the AguaFunda Park: P7, under the trees, and P5, in

the park gate. In the case of P7, due to thedense shading, as one may observe in Figure3, there is a smaller sky view factor, resultingin neutrality conditions during the period. AtP5, the warm condition continues until10:45am. In this case, the less severe thermalcondition is achieved due to a relative smallersky view factor and to the proximity to thedense vegetation of the park.

The third period of measurements,from 1 pm to 2 pm presents very hotsituations in all the points of study, except forP7, in which one may find hot situations. Thispoint presented the better performance mainlydue to the shading from the trees, not onlyblocking the direct solar radiation but also

providing air temperature stability.

During the last period ofmeasurements, from 4pm to 5 pm, one mayobserve again the predominance of very hotsituations, except for P7, where the attenuationfound is explained for the reasons presented inthe last paragraph. The different results foundat P2 and P1, where attenuation is foundrespectively at 4:30pm and 4:45pm, may beexplained due to the fact that these points arerelatively far from the others, being located in a

more uniform horizontal urban tissue. Theother points are located in less built areas, butmore arid ones, due to pavements used andabsence of vegetation (P3 e P4) or inside thepark (P5, P6 e P7).

The differences in the results of P1and P2 can be also explained due to theweather changes in the last period ofmeasurements, as one may concludeobserving the results from the meteorologicalstation. Maybe the results of P1 and P2 couldbe explained because they have a greater skyview factor, contrarily to the park, which wasprotect by dense vegetation. P3 is also wellexposed, but the same difference was notfound in this point, following the results of P4,P5 e P6, which are not so exposed. One mustmention that measurements were previewed totake place also between 7pm and 8pm, butthey were not performed due to the climaticinstability. These observations take intoaccount the plausible consideration that, basedon the registered data, the instability arrivedfirst at the meteorological station, then to thestudied points.

8/10/2019 126712

7/8

Considering now specifically theresults found based on the data from themeteorological station, one may observe that,during the first period of measurements,between 7am and 8am, there are neutralitythermal sensations, in concordance with the

results found on the field. This shows that, inthe absence of solar radiation, the results frommeteorological station seem to be veryrepresentative of the general conditions foundin the different studied points.

During the second period ofmeasurements, from 10am to 11 am, one mayobserve hot situations, also in accordancewith the general results from the field. P5 andP7 are exceptions, as they present smaller skyview factors, respectively because there are

shading by high rise buildings and by densetrees. It must be mentioned that, during thefirst measurement of the second period, thepredominant situation was warm, whilst theresults from the meteorological stationsshowed a hot situation. This can be explainedfor the fact that all the points present somekind of blockage to the solar radiation duringthe first hours of the morning, whilst themeteorological station is situated in anunobstructed location. This consideration iscorroborated by the results from P3, which

presents, during the first measurement of thesecond period, a situation of hot, explainedby the fact that there is practically noobstruction in its surroundings.

In relation to the third period ofmeasurements, between 1 pm and 2 pm, thereare two situation of very hot followed by threeof hot. Except for P7, which presents denseshading from the trees, all the other studiedpoints presented very hot situations duringthe completely third period. Two questions

should be considered. Firstly, as it was saidearlier in this work, the climatic instabilityarrived first at the meteorological station.Secondly, despite the fact that Barra Funda isnot a dense built area, it is almost completelypaved, presenting considerable thermalcapacity, lengthening the period of very hotsituation.

Finally, during the last period ofmeasurements, from 4 pm to 5 pm, one mayobserve that the first result from the

meteorological station is a warm condition,followed by neutrality ones. One may verifyhere the same tendency pointed out in the last

paragraph. If it was possible to perform themeasurements from 7 pm to 8 pm, probablyresults of hot, warm and neutral conditionswould be consecutively observed in the variouspoints of study, with different timediscrepancies due to the geographical location

and local thermal inertia from the builtenvironment.

6. FINAL CONSIDERATIONS

Differences in the final results of thestudied points were observed, pointing out topossible relationships between urbanmorphologies and vegetation and the localmicroclimate, and also to differentconsiderations concerning the environmentaladequacy in terms of thermal sensation of theacclimatized population. The results found andthe considerations made show the relevance ofregarding the urban design and the localtreatment of open spaces in order to constituteoutdoor spaces that are thermally proper,allowing conditions to the effective use of theurban spaces.

Considering the hot climate, highdensity constructions showed to allow longerperiods in comfort conditions, due to the

decrease of the sky view factor andconsequently the significant shading effectupon the area. As expected, green areaspresented even better results, considering theshading and the evapotranspiration thatinfluences in the thermal comfort sensation. Inhot urban environments, the use of higherdensities and green areas seems to providegood results not only in the streets, but alsowhen configured as pocket parks, which takeuse of the advantages of both strategies.

This ongoing research has two maindevelopment lines, which will be object offuture publications. The first one details thethermal analysis done by means of: theverification between the original and thecalibrated indexes, the consideration in fulldetail of the influence of the radiation andventilation in the microclimatic conditions, andthe comparative application of an adaptivemodel. The second one tests the results foundproposing urban interventions, contributing topossible urban redesigns and to the properdesign of empty areas in the surroundings ofthe study area.

8/10/2019 126712

8/8

7. REFERENCES

ASHRAE. 2001 Fundamentals Handbook.American Society of Heating, Ventilating andAir-Conditioning Engineers. Atlanta. USA.

BLAZEJCZYK, Krysztof. 2002 Assessment ofrecreational potentional of bioclimate. in:International Workshop on Climate, Tourismand Recreation, 1, 2001, Greece. ISB, p. 133-52.

______. 2004 Man-environment heatexchange model. Accessed in 24/04/2004.http://www.igipz.pan.pl/klimat/blaz/menex.ppt.

DUARTE, D; GONALVES, J. 2006Environment and Urbanization: MicroclimaticVariations in a Brownfield Site in Sao Paulo,Brazil. The 23rd Conference on Passive and

Low Energy Architecture. In: ProceedingsGeneva: PLEA, September 2006.

ISO 1998. ISO 7726. Ergonomics:instruments for measuring quantities. ISO,Genve, Switzerland.

______. 1995. ISO 9920. Ergonomics of thethermal environment: estimation of the thermalinsulation and evaporative resistance of aclothing ensemble. ISO, Genve, Switzerland.

______. 1990. ISO 8996. Ergonomics:metabolic heat production. ISO, Genve,

Switzerland.

MONTEIRO, L.M.; ALUCCI, M.P. 2007Conforto trmico em espaos abertos comdiferentes abrangncias microclimticas. Parte1 e 2. Encontro Nacional de Conforto noAmbiente Construdo. In: Anais... Macei:ENCAC, agosto de 2007.

______. 2006. Verificao comparativaexperimental da aplicabilidade de diferentesmodelos preditivos de conforto trmico.Encontro Nacional de Tecnologia do Ambiente

Construdo. In: Anais...Florianpolis: ANTAC,agosto de 2006.

So Paulo 1995. Prefeitura Municipal de SoPaulo. Lei 11.774/95. Operao Urbanagua Branca.

8. ACKNOWLEDGEMENT

We would like to thank all the graduatestudents that have helped in the empiricalresearch: Anarrita Buoro, Anderson Nedel,Anna Miana, Antonio Machado, Carolina Leite,Daniel Costola, Erica Umakoshi, Jos Ramos,Juliano Beraldo, Marcos Yamanaka, MonicaMarcondes, Paula Shinzato, Rafael Brandaoand Sabrina Agostini. We are also thankful toProf. Koen Stemeers of The Martin Centre forArchitectural and Urban Studies, University ofCambridge, and to Prof. Susannah Hagan, ofUniversity of East London, for the discussionthat preceded this research in meetingsfinanced by The British Academy. Finally, wewould like to thank the Fundacao de Amparo a

Pesquisa do Estado de Sao Paulo for thefinancial support.