A Numerical Study of Natural Ventilation in Buildings - Utilized Solar Chimney and Cool Ceiling

Transcription

1 The 2 nd Jont Internatonal Conference on Sustanable Energy and Envronment (SEE 2006) B-043 (P) November 2006, Bangkok, Thaland A Numercal Study of Natural Ventlaton n Buldngs - Utlzed Solar Chmney and Cool Celng Sudaporn Chungloo and Bundt Lmmeechokcha * Srndhorn Internatonal Insttute of Technology, Thammasat Unversty, P.O. Box 22 Thammasat Rangst Post Offce 12120, Pathumtha, Thaland Abstract: The objectve of ths paper s to nvestgate the beneft of applcaton of solar chmney on the south roof and water sprayng on the north metal celng of a test buldng through a numercal model constructed from a computatonal flud dynamcs (CFD) program. The two dmensonal numercal model are generated by the CFD model usng the expermental data of walls and ar temperatures durng applcaton of solar chmney n a test buldng as the boundary condtons. The predcted results are compared wth the selected expermental results before the temperature of 2 C and 4 C of the cool celng are appled to the valdated model n order to study the effect of cool celng on the temperature, velocty and ar flow rate n the test buldng. The smulaton results show a satsfactory agreement wth the expermental results for the temperature dstrbuton n the test buldng. The dfferences between numercal results and the expermental results of volume flow rate are 13-18%. Accordng to the numercal smulaton, the reducton of temperature of the north celng by 2 C and 4 C ncreases the volume flow rate at the ext of the solar chmney by 6%and 17%, respectvely. Keywords: Passve Coolng n Buldng, Natural Ventlaton, Solar Chmney, Water Sprayng, Smulaton 1. INTRODUCTION As a devce for natural ventlaton, a solar chmney has been found as an effectve passve coolng desgn n the hot and humd clmate [1-3]. An approprate desgn of a solar chmney for natural ventlaton ncludes provdng an ar gap n a south façade or n the roof of the buldng that causes stack effect between the solar chmney and the nlet of the buldng. The stack effect operates between the hgh temperature and hgh pressure developed n the solar chmney and the low pressure and low temperature at the nlet. If the openngs are provded at the nlet of the buldng and at the outlet of the solar chmney, ar wll enter nto the buldng due to the dfference of ar denstes and pressure gradent and move through the buldng before leavng the outlet of the solar chmney. Therefore, the performance to produce a certan ventlaton flow rate of a solar chmney depends on the temperature dfference between the nlet and the outlet. In addton, the hgh ambent temperature, hgh wnd velocty and hghly fluctuatng of solar radaton n hot and humd clmate can effect the ventlaton flow rate of a solar chmney. There are some modfcatons of the solar chmney for ncreasng the ar flow rate through buldngs [4-7], and utlzng daylghtng [8]. The recent study of reducton of temperature by water sprayng on the celng of a test buldng for ncreasng the temperature dfference between the nlet of the buldng and outlet of the solar chmney found that the temperature of the ar n the test buldng was reduced by 1.4- C comparng to the buldng wthout water sprayng [9]. The purpose of ths paper s to numercally nvestgate the effect of applcaton of solar chmney and water sprayng on roof on the temperature feld, velocty feld, and volume flow rate of the test buldng. 2. METHODOLOGY 2.1 Test buldng The test buldng s a detached one-story room located on the roof of Srndhorn Internatonal Insttute of Technology (SIIT) s laboratory buldng, Thammasat Unversty. The nteror dmensons are 3.8m 2.8m 2.4m. Its walls are consttuted by wood frame externally covered wth whte-panted asbestos-cement board and nternally covered wth plywood board. A layer of polystyrene foam of 76 m s located n the wood frame between the asbestos-cement board and the plywood. The ptch roof wth tlt angle of 45, composes of terracotta roof tle at the outer sde, 0.12 m ar gap and the gypsum board at the nner sde. The ar gap under the south roof s the solar chmney where the outdoor ar flowng through the wndow of the room reduces hot ar storage n ths gap (Fg 1 and 2). The test buldng s placed at the poston that each of the four wall facng to east, west, north and south drectons. 2.2 Testng methodology The experments were carred out on selected days durng May-June Fgure 2 shows that total 16 thermocouples of type K connected to the Yokogawa-DR130 data logger were used for temperature measurement at varous ponts n the test buldng. A wnd montorng devce s connected to the Yokogawa-DR130 data logger to collect wnd velocty. A pyranometer, Kpp&Zonen-CM11, was connected to Yokogawa data logger to collect global solar radaton n W/m 2. Two hot-bulb velocty probes were connected to the Testo-454 data logger to record velocty at the nlet and outlet of the solar chmney. All data were recorded every 2 mnutes for 24 hours. Due to the hgh fluctuaton of the solar radaton, walls/roof temperatures, ambent temperature and ndoor ar temperature, t s very dffcult to obtan a steady condton n the test buldng. Hence, the quas-steady condtons are assumed when the solar radaton dd not fluctuate more than 100 W/m 2 and the temperature of the roof surface dd not change more than C. Before expermental startup all temperature probes were calbrated wth a hgh accuracy and calbrated thermometer. Fg. 2 and Table 1 show the postons of measurement and the average values of the four selected expermental results, respectvely. These expermental results are used for boundares and the valdaton of the result from smulaton. Correspondng author: bundt@st.tu.ac.th 1

2 The 2 nd Jont Internatonal Conference on Sustanable Energy and Envronment (SEE 2006) B-043 (P) November 2006, Bangkok, Thaland Fg. 1 The test buldng Fg 2. Postons of the ar temperature measurements Table 1 The average values of selected expermental results Case#1 Case#2 Case#3 Case#4 Secton # Indoor surface temperature ( C) Roof tle Gypsum board Celng South wall North wall Floor Ar temperature ( C) Outflow from solar chmney Room temperature Inflow to room Ambent Velocty at nflow to solar chmney (m/s) Velocty at outflow from solar chmney (m/s) Wnd velocty (m/s) Solar radaton (W/m 2 ) The numercal model A commercal CFD package, FLUENT [10], s used for smulaton the two dmensonal model of the test room. The two dmensonal lamnar and low Renolds number turbulent flow modelng s an mportant study of room arflows [11]. Hence, the low turbulent flow under the buoyancy effect n the test room and ts solar chmney were carred out by usng the Renormalsaton Group (RNG) k-ε model and the Boussnesq s approxmaton provded by the FLUENT package, respectvely. The followng basc conservaton equatons are solved by the FLUENT program. Contnuty: + ( ρu ) = 0, ρ t x (1) p τ j t x j x x (2) j T p p u ρ t x x x t x j (3) x j Momentum: ( ρ u ) + ρu u = + + ρg, Energy: ( h) + ( ρu h) = k + + u + τ, where ρ s the densty, u s the velocty component n the -drecton, p s the statc pressure, t s the tme, x s a Cartesan coordnate, τ j s the stress tensor, g s the gravtatonal acceleraton n -drecton, h s the statc enthalpy, k s the thermal conductvty, and T s the temperature. Snce the flow s the buoyancy-drven, Eq. (2) the Boussnesq s model s used for the vertcal x 2 -drecton as follows: Boussnesq s approxmaton: ( ρ ρ0) g ρ0β ( T T0 ) g, (4) where ρ 0 s a constant densty of the flow, T 0 s the operatng temperature and β s the thermal expanson coeffcent. The boundary condtons for the momentum equaton are no-penetraton, no-slp at all the boundares and p=0 s mposed at the openngs. The boundary condtons for the energy equatons are based on the constant wall temperatures found durng the quas-steady state n the 2

3 The 2 nd Jont Internatonal Conference on Sustanable Energy and Envronment (SEE 2006) B-043 (P) November 2006, Bangkok, Thaland experments. The room and the solar chmney are smulated together as the same doman where t s dvded nto 4,284 rectangular grd cells where fner grd sze s used near the wall regons. 3. RESULTS AND DISCUSSION 3.1 Temperatures at varous heghts n the test buldng The comparsons between the expermental and numercal results of temperature dstrbuton along the heght of the test buldng are shown n Fg. 3. All results obtaned from the smulatons and the experments show the thermal stratfcaton n the test buldng. In case #1, 2 and 3, good agreements between the smulaton results and expermental results are found at the mddle heght of the room (between 0.8 and m) where the man stream flows from the wndow on the north wall to the solar chmney near the south wall. However, the temperature of ar near the floor and near the celng tends to be under-predcted by the numercal smulaton n case # 1, 2 and 3. The hgh temperature of ar near the floor and near the celng s the result of recrculaton of the man stream of hgh temperature whch s not expelled from the room through the solar chmney. Case # 4 (secton 2 n the room) shows the effect of postons of measurement and low thermal buoyancy force on the temperature dstrbuton n the room. At secton 2 of the room, the velocty of ar at the nflow to the solar chmney s low because the man stream of ar does not flow to ths regon and the dfference between the temperature of ar n the room and at the outflow from the solar chmney n case #4 s low, comparng to case #2. In addton, the low temperature at surfaces of nteror walls n the test buldng and no recrculaton of the man stream keeps temperature of ar n ths regon lower than the temperature n secton 1 and 3. At the heght of solar chmney between and 3.9 m, ar temperatures from smulaton results are lower than ar temperature from expermental results specfcally at the nflow to the solar chmney. Ths can be the effect of the under-predcton of the ar temperature n the room on the predcted values of ar temperature n the solar chmney. However, the trends of the thermal stratfcaton are satsfactory for case #1 and #3, and acceptable for case #2 and # C 37.0 C 39.0 C 38.7 C Exp erment (solar chmney) Predct (solar chmney) Experment (Room) Predct (Room) C 37.4 C 37.2 C 36.8 C Exp erment (Roo m) Predct (Room) Case #1 Case # C 33.2 C 34.2 C 34.8 C Experment (solar chmney) Predct (solar chmney) Exp (Rm) Predct (Rm) Experment (solar chmney) Predct (solar chmney) 3 C 29.8 C 29.7 C 29.4 C Exp erment (so lar chmney) Predct (solar chmney) Experment (Room) Predct (Room) Case #3 Case #4 Fg. 3 Comparson between expermental and smulaton results of ar temperature n the test room and n the solar chmney 3.2 Volume flow rate per unt wdth of solar chmney The measured values of velocty and temperature at the outflow from the solar chmney were used for calculaton of the total volume flow rate of the solar chmney. The derved values of total volume flow rate were dvded by the total wdth of the solar chmney, whch s 0.95 m 4 = 3.8 m, before comparng wth the volume flow rate computed from the numercal smulaton. The comparsons between the expermental and numercal results of volume flow rate per unt wdth of the test buldng are shown n Table 2. A certan agreement between the smulaton results and expermental results are found n case #1 and case #3 wth percent dfferences of 13.1 and 18.3%, respectvely. The hgh fluctuaton of measured veloctes durng the experment causes the dfference between the expermental and the numercal results. However, the error wthn ±20% was accepted here [12]. For case #2 and case #4, the relatve errors are more than 25%. The results from numercal smulaton of case # 1, #2 and #3 are used for the study of the effect of reducton of temperature on the north celng by 2-4 C. The results of smulaton of temperature reducton at the celng are also shown n Table 2. The volume flow 3

4 The 2 nd Jont Internatonal Conference on Sustanable Energy and Envronment (SEE 2006) B-043 (P) November 2006, Bangkok, Thaland rate per unt wdth of the solar chmney ncreases by 5.9, 10.1 and 7.8%, for celng temperature reducton of 2 C. The further reducton of celng temperature to 4 C shows that the volume flow rate per unt wdth of the solar chmney ncreases by 10.9, 16.7 and 1% for the study cases # 1, #2 and #3, respectvely. Table 2 Volume flow rate per unt wdth of solar chmney Case##1 Case#2 Case#3 Case#4 Applcaton of solar chmney Expermental result (m 3 /s m) Smulaton results (m 3 /s m) (%) relatve error Applcaton of solar chmney wth cool roof Predcted values (m 3 /s m) after reduce temperature at north celng by 2 C Predcted values (m 3 /s m) after reduce temperature at north celng by 4 C % ncrease (for reduce 2 C) % ncrease (for reduce 4 C) Numercal study of the solar chmney and the effect of cool roof on the temperature and velocty feld The numercal study usng a computatonal flud dynamc (CFD) program s an effectve way to nvestgate the temperature and velocty felds n the test and real buldngs. In ths study, the transent smulatons wth the tme step of 0.1 sec. were performed for all four cases and the steady state were approached after 20 mnutes. The transent processes of all cases are qute smlar, so the process of case #1 at 1 sec., 6 sec. and 10 mnutes are shown n Fg 4. Fg. 4 Temperature dstrbuton (upper row) and velocty feld (lower row) durng the transent processes of case #1 wth solar chmney The process starts from the ambent ar enter the wndow wth the partcular mushroom-lke shape descendng to the floor [13] at 1 sec. At 6 sec., the mushroom-lke shape grows and dsperses n the low regon of the room where the regon of low temperature under the enterng stream and the regon of the mxng of ar at varous temperatures above the enterng ar stream are generated. The correspondng velocty felds n Fg. 4 show the clockwse recrculaton and the counter-clockwse rotaton of ar under the enterng stream and above the enterng stream, respectvely. The processes of mxng of ar streams contnue untl the enterng stream lft to 4

5 The 2 nd Jont Internatonal Conference on Sustanable Energy and Envronment (SEE 2006) B-043 (P) November 2006, Bangkok, Thaland the upper regon of the room at 10 mnutes. Here, the enterng ar stream s dvded nto two parts at the mddle of the room: one part flows to the solar chmney and another part re-crculates n the regon near the wndow. The recrculaton of ar can also be found n the solar chmney near the gypsum board because the large heat loss from the hot ar to the gypsum board [14]. Fgure 5a and 5b show the smulaton results of temperature dstrbutons and velocty felds at the steady condton of the case #1 wth and wthout applcaton of the cool celng, respectvely. There are two stably stratfed regons n both Fg 5a and 5b, n the solar chmney and n the low part of the room. The cool floor s below the warm prmary stream, therefore the recrculaton s suppressed and the thermal stratfcaton s generated [15]. In the upper part of the room, there s an exstence of the nearly unform temperature for the applcaton of solar chmney wthout cool celng (Fg 5a) but there are two regons wth dfferent temperatures appear n the upper part of the room n the applcaton wth cool celng (Fg 5b). The temperature dfference between the warm enterng stream and the cool celng creates the buoyancy effect n the upper regon of the room. The ar entraned by the warm jet and the cool celng has dfference densty and t begns to move n the counterclockwse drecton. Accordngly, the velocty feld n Fg. 5b shows a well-defned crculatng at the upper regon of the room and the descendng of hgh densty ar above the nlet of the room. The crculaton of ar n the case wthout cool celng s also appeared but t s weak comparng to the case wth cool celng. Therefore, The smulaton result show that the applcaton of cool celng together wth the solar chmney reduce temperature of room ar, ncrease the velocty and volume flow rate of ar n the buldng. (a) (b) Fg. 5 Temperature and velocty felds from smulaton (a) solar chmney (b) solar chmney wth celng temperature reducton of 2 C 5

6 The 2 nd Jont Internatonal Conference on Sustanable Energy and Envronment (SEE 2006) B-043 (P) November 2006, Bangkok, Thaland 4. CONCLUSIONS The numercal study usng a computatonal flud dynamc (CFD) program s an effectve way to nvestgate the temperature dstrbutons and velocty felds n the buldngs. In ths study, t s observed that a satsfactory agreement between the smulaton and expermental results can be acheved by usng a smplfed two-dmensonal CFD model wth constant wall temperatures. The results from smulaton show that the cool celng applcaton together wth the solar chmney on roof can ncrease the velocty n the room and enhance the volume flow rate of ar through the solar chmney from 6-17% for the reducton of celng temperature of 2-4 C. Accordng to ths smplfed model, the temperature of ar n the room s reduced by C, specfcally n the regon under the cool celng. 5. ACKNOWLEDGEMENTS The authors would lke to acknowledge the support of the CFD program at the Mechancal Engneerng Deartment, Kng Mongkut s Insttute of Technology North Bangkok (KMITNB). In addton, the authors would lke to thank the Thaland Research Fund (TRF) for the fnancal support on ths research, and Prof. Vmolsddh Horayanggura from the Faculty of Archtecture, Thammasat Unversty for the comments and desgn of the test buldng. 6. REFERENCES [1] Bansal, N.K., Mathur, R., Bhandar, M.S. (1993) Solar chmney for enhance stack ventlaton, Buldng and Envronment (28), pp [2] Bansal, N.K., Mathur, R., Bhandar, M.S. (1994) A study of solar chmney asssted wnd tower system for natural ventlaton n buldngs, Buldng and Envronment (29), pp [3] Khedar, J., Mansrsub, W., Chama, S., Prathnthong, N., Hrunlabh, J. (2000) Feld measurements of performance of roof solar collector, Energy and Buldngs (31), pp [4] Aboulnaga, M.M. (1998) A roof solar chmney asssted by coolng cavty for natural ventlaton n buldngs n hot ard clmates: an energy consderaton approach n Al-An cty, Renewable Energy (14), pp [5] Khedar, J., Boonsr, B., Hrunlabh, J. (2000) Ventlaton mpact of a solar chmney on ndoor temperature fluctuaton and ar change n a school buldng, Energy and Buldngs (32), pp [6] Hrunlabh, J. Wachrapuwadon, S., Pratnthong, N. Khedar, J. (2001) New confguratons of a roof solar collector maxmzng natural ventlaton, Buldng and Envronment (36), pp [7] Gan, G. and Rffat, S.B. (1998) A numercal study of solar chmney for natural ventlaton of buldngs wth heat recovery, Appled Thermal Engneerng, (18), pp [8] Waewsak, Hrunlabh, J., Khedar, J. and Shn, U.C. (2003) Performance evaluaton of the BSRC mult-purpose bo-clmatc roof, Buldng and Envronment, (38), pp [9] Chungloo, S. and Lmmeechokcha, B. (2006) The Feld Investgaton On The Applcaton of Solar Chmney and Wetted Roof n the Hot and Humd Clmate. World Renewable Energy Congress IX, LEA 58, Florence Italy [10] Fluent User s Gude, (snce 1995) Fluent Inc, USA. [11] Snha, S.L., Arora, R.C. and Roy, S. (2000) Numercal smulaton of two-dmensonal room ar flow wth and wthout buoyancy. Energy and Buldngs (32), pp [12] Barozz, G.S., Imbab, M.S.E. and Sousa, A.C.M. (1992) Physcal and numercal modellng of a solar chmney-based ventlaton system for buldngs, Buldng and Envronment (27), pp [13] Dubovsky, V., Zsknd, G., Druckman, S., Moshka, Y. and Letan R. (2001) Natural convecton nsde ventlated enclosure heated by downward-facng plate: experments and numercal smulatons, Internatonal Journal of Heat and Mass Transfer (44), pp [14] Gan, G. (2006) Smulaton of buoyancy-nduced flow n open cavtes for natural ventlaton, Energy and Buldng (38), pp [15] Snha, S.L., Arora, R.C., Subhransu, R. (2000) Numercal smulaton of two-dmensonal room ar flow wth and wthout buoyancy, Energy and Buldng (32), pp