Numerical investigations of buoyancy-driven natural ventilation in a simple three-storey atrium building and thermal comfort evaluation

Applied Thermal Engineering 57 (2013) 133e146

Contents lists available at SciVerse ScienceDirect

Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng

Numerical investigations of buoyancy-driven natural ventilation in a simple three-storey atrium building and thermal comfort evaluation Shafqat Hussain, Patrick H. Oosthuizen* Department of Mechanical and Materials Engineering, Queen’s University, 130 Stuart Street, Kingston, ON, Canada K7L3N6

h i g h l i g h t s
A simple three-storey atrium building.
Numerical modeling of buoyancy-driven ventilation flow in the building.
Effect of solar intensity and geographical location on ventilation.
CFD predictions were used to calculate thermal comfort indices.
Evaluation of thermal comfort conditions for the occupants.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 30 October 2012 Accepted 19 March 2013 Available online 27 March 2013

The numerical investigations of buoyancy-driven natural ventilation and thermal comfort evaluation in a simple three-storey atrium building as a part of the passive ventilation strategy was undertaken using a validated Computational Fluid Dynamic (CFD) model. The Reynolds Averaged NaviereStokes (RANS) modeling approach with the SST-keu turbulence model and the discrete transfer radiation model (DTRM) was used for the numerical investigations. The steady-state governing equations were solved using a commercial solver FLUENTÓ. Various flow situations of the buoyancy-driven natural ventilation in the building during day and night time were examined. The numerical results obtained for the airflow rates, airflow patterns and temperature distributions inside the building are presented in this paper. Using the numerical results, the well-known thermal comfort indices PMV (predicted mean vote) and PPD (predicted percentage of dissatisfied) were calculated for the evaluation of the thermal comfort conditions in the occupied regions of the building. It was noticed that thermal conditions prevailing in the occupied areas of the building as a result of using the buoyancy-driven ventilation were mostly in comfort zone. From the study of the night time ventilation, it was found that hot water (80  C) circulation (heated by solar collectors during daytime) along the chimney walls during night time and heat sources present in the building can be useful in inducing night ventilation airflows in the building as a part of the passive ventilation strategy. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Atrium building Night ventilation Numerical modeling Thermal comfort evaluation

1. Introduction Low-energy cooling technologies provide cooling in an energyefficient manner, thus reducing building energy consumption and peak electricity demand. They do so by making use of low quality sources of cooling whether it is ambient air or ground temperatures or warmer chilled water. These technologies may be considered as passive and hybrid cooling systems. The basic method to design and analyze a building with low-energy cooling technologies is building

* Corresponding author. Tel.: þ1 613 533 2573. E-mail address: [email protected] (P.H. Oosthuizen). 1359-4311/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.applthermaleng.2013.03.033

simulation. In these applications the dynamics and interactions of building, systems, occupants and environment are complex and very important. Traditional design methods assume only the extreme boundary conditions, and therefore they are not suitable for the design and analysis of low-energy cooling systems. The computer based modeling and simulation approaches are more cost effective, less time-consuming and more reliable for the numerical investigations to analyze the indoor thermal environment of buildings. Several studies have shown night ventilation to be effective in improving thermal comfort and reducing cooling demand during daytime, e.g., see Refs. [1e11]. Yang and Li [12] studied the effect of the thermal mass on energy consumption in office buildings with

134

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146

air-conditioning at daytime and free cooling at night time to quantify the hourly and overall variation of cooling load of airconditioning. It was found that an increase of time constant could effectively reduce the cooling load. Geros et al. [13] investigated the potential of night ventilation strategies applied to full-scale buildings under different structure, design, ventilation and climatic characteristics using both experimental and simulation tools. It was shown that the efficiency of night ventilation was strongly related to three main parameters: the relative difference between indoor and outdoor temperature mainly during night, the useful airflow rate applied during the night period, and thermal capacity of the building. It was noted that night ventilation techniques could contribute to decrease significantly the cooling load of air-conditioning and improve the comfort levels of buildings. The exact contribution of night ventilation for a specific building is a function of the building structure and design characteristics, the climatic conditions and the building layout, the applied airflow rate, the efficient coupling of airflow with thermal mass of the building, and the assumed operational conditions. Santamouris et al. [14] presented the analysis of the energy data from two hundred fourteen air-conditioned residential buildings using night ventilation techniques and calculated the specific energy contribution of night ventilation to find out the relation with the cooling demand of the building. It was found that global utilization of the energy stored during night time increased as a function of the airflow rate. Pfafferott et al. [15] studied the design of passive cooling by night ventilation and evaluated building simulations with measurements at the new institute building of Fraunhofer ISE in order to determine the efficiency of night ventilation dependant on air change rate, solar and internal heat gains. It was concluded that the building simulations provide accurate results, if the input parameters and boundary conditions are well known. Geros et al. [4] conducted an experimental evaluation of night ventilation in four different buildings and used numerical investigations to determine how the air change rates, the building construction and the climatic parameters affect the night ventilation. Letan et al. [16] carried out experimental and numerical study of passive ventilation and heating in a multi-storey structure by natural convection in a heated vertical duct. This study showed that the results obtained from simulations supported by the measurements indicated that effective ventilation and heating were achievable in the laboratory structure and in the real-size five-storey building. It was noted that even at low solar irradiation fluxes ventilation was achieved in summer, and heating in winter. Bansal et al. [17] used a solar chimney to enhance stack ventilation and achieved the flow rates of hundreds of cubic meters of air per hour. Givoni [18] performed experiments in a low and a high mass building in order to determine the effectiveness of mass and night ventilation in lowering the indoor daytime temperature. Afonso and Oliveira [19] conducted simulations and experiments on a glazed solar chimney constructed on the southern wall of a building. From literature survey it was noted that night ventilation has been applied successfully in many passively cooled or low-energy office buildings. Night ventilation has been proved to be an effective cooling method, especially in regions with a large temperature difference between day and night. To create a comfortable indoor climate during summer, the building must be cooled at night in order to achieve the correct temperature at the start of the following working day. Night ventilation of buildings can be incorporated as part of a passive ventilation strategy to purge the building of warm air that has accumulated during the day and, thereby, to cool the building fabric in order that it may act as a heat sink during the following day. The application of computational fluid dynamics (CFD) to evaluate the thermal comfort conditions in atrium buildings has been

studied by some researchers, i.e., see Refs. [20e23]. Standards and fundamentals of thermal comfort have been published, e.g., see Refs. [24e28]. Thermal comfort is defined in ISO 7730 [28] as “the condition of mind that expresses satisfaction with the thermal environment”. In recent years, the most common thermal comfort model used is the traditional predicted mean vote (PMV) index model for air-conditioned buildings [27]. The thermal comfort indices PMV (predicted mean vote) and PPD (predicted percentage of dissatisfied) of Fanger and Toftum [27] are accepted as ISO 7730 and are calculated by empirical equations. The PMV equation is based on steady-state heat balance for the human body and is expressed as follows [27].

PMVtrad ¼ ½0:303expð0:036MÞ þ 0:028L

(1)

where PMVtrad is the traditional PMV (for air-conditioned buildings), M is the metabolic rate (W/m2) and L is the thermal load on the body expressed as follows: L ¼ internal heat production  heat loss to the actual environment

h i n L ¼ M  W  3:96  108 fcl ðTcl þ 273Þ4  ðTr þ 273Þ4 þ fcl hc ðTcl  TÞ þ 3:05  103 ½5733  6:99ðM  WÞ  Pv  þ 0:42ðM  W  58:15Þ þ 1:7  105 Mð5867  Pv Þ o þ 0:0014Mð34  TÞ (2) where W stands for active work or shivering (W/m2) and fcl is the garment insulation factor (1 clo ¼ 0.155 m2 K/W) expressed as

fcl ¼ f1:05 þ 0:645Icl

for Icl

> 0:078 and 1 þ 1:291Icl

for Icl < 0:078g

(3)

The term Icl stands for the resistance to sensible heat transfer provided by a clothing ensemble (clo). The Tcl ( C) term is defined as the cloth temperature and is determined below as:

n Tcl ¼ 35:7  0:028ðM  WÞ  Icl 3:96  10 h i o  108 fcl ðTcl þ 273Þ4  ðTr þ 273Þ4 þ fcl hc ðTcl  TÞ (4) ( C)

In Eqs. (2) and (4), T is local air temperature, hc, is the heat transfer coefficient between the cloth and air (W/m2 k) and Tr ( C) is the mean radiant temperature. The heat transfer coefficient is given by:

hc ¼ 2:38ðTcl  TÞ0:25; > 12:1u0:5

and

for 2:38ðTcl  TÞ0:25

hc ¼ 12:1u0:5 ;

for 2:38ðTcl  TÞ0:25

0:5

< 12:1u

(5) where u is the local velocity. However, it has been proven that the traditional PMV index is inadequate for naturally ventilated buildings, and an optional thermal comfort model, also known as ‘‘adaptive comfort standard’’ is implemented in the new revised American Society of Heating, Refrigerating and Air-conditioning Engineers (ASHRAE) Standard 55. This model is used to determine thermal comfort in exclusively naturally ventilated spaces. It has limitations [27] such as that it is only applicable to mean monthly temperatures from 10  C to 33  C

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146

and that only information of global thermal discomfort in large spaces is provided. Fanger and Toftum [27] introduced an extended PMVNV (natural ventilation PMV) comfort model more suitable for naturally ventilated buildings which can be considered consistent with the traditional PMV model. The PMVNV index represents an extension of the traditional PMV model to account for the occupants ‘expectancy factor’ according to their habitat, as well as to the estimated activity, i.e., the metabolic rate is reduced for every scale unit of PMV above neutral, a PMV of 1.5 corresponds to a reduction in the metabolic rate of 10% under hot and humid conditions. The expectancy factor is multiplied by the traditional PMV to give the mean thermal sensation vote of the occupants of a naturally ventilated building in a warm climate. The expectancy factor is estimated to vary between 1 and 0.5. If the weather is warm all year or most of the year, expectancy factor may be 0.5. In regions with only brief periods of warm weather occur during the summer, the expectancy factor may be 0.9e1.0. It is 1.0 for the air-conditioned buildings. In this model, PMV and PPD (%) for natural ventilation are determined as follows:

PMVNV ¼ e½0:303expð0:036Mred Þ þ 0:028L

(6)

h PDð%Þ ¼ 100  95exp  0:03353ðPMVNV Þ4 i  0:2179ðPMVNV Þ2

(7)

where, e is the expectancy factor and Mred is reduced metabolic rate. The PPD equation indicates the variance in the thermal sensation of a group of persons exposed to the same thermal conditions. Dissatisfaction with the thermal environment, discomfort was defined by participants using the 7-point scale: cool (2), cold (3), warm (þ2) or hot (þ3). Under optimal thermal conditions (PMV ¼ 0), 5% of persons will be dissatisfied assuming identical activity level, clothing, and environmental conditions. The PMV and PPD are calculated from six basic variables: activity, clothing, air temperature, air velocity, mean radiant temperature (MRT) and humidity. The values of the variables for the activity (metabolic rate) and clothing (ensemble insulation) are determined using ASHRAE Fundamentals. In order to quantify thermal comfort in naturally ventilated buildings, these models may be integrated with a CFD model. The aim of the present research was to explore the use of buoyancy-driven ventilation and evaluate the thermal conditions in a simple three-storey full-scale atrium building integrated with a combination of atrium and chimney. This study is an additional step of the numerical investigations conducted in our laboratory during the recent years on different aspects of natural convection and buoyancy-driven ventilation where the numerical techniques used have been validated against the experimental measurements available, e.g., see Refs. [29e34]. The model used captures how the pressure differences produced by warm air that has accumulated in each storey and atrium drive a ventilating flow through each storey and discharge into the atrium as turbulent thermal plumes. The model predicts the depth and temperature of the layers of warm air in each storey and the atrium.

Refs. [29e34]. Based on the analysis of the effects of changes in various geometric and climatic parameters on the performance of the building presented in an another paper [34] and keeping in mind the possible use of night time natural (buoyancy-driven) ventilation and heat losses during winter, a few changes in the selected geometry were carried out to examine their effect on the ventilation airflow, temperature distributions and thermal comfort conditions in the building. The selected geometry of the simple atrium building was considered with some changes in its design (i.e., (i) atrium integrated with a conventional chimney having a central heated plate 3 m high, (ii) chimney height increased from 4 m to 6 m, (iii) height of inlets above floors on ground and first floors increased to 1.1 m, and (iv) façade glazing area reduced from 94 to 60 m2) shown in Fig. 1. The important dimensions and areas of the building are given in Table 1. The values of the climatic parameters used in the CFD simulations are shown in Table 2. In order to utilize the validated CFD model [29e33], the simulated building was assumed to be located and oriented like the Engineering building at Concordia University, Montreal, Canada i.e., 35 west of south, with the same kind of façade glazing surface facing southwest. The atrium exhausts were located on the highest point in the atrium and storey inlets and outlets on each storey were located on the side walls of the rooms (Fig. 1). The inlets and outlets were sized on each storey using the design curves developed by Holford and Hunt [35]. The effective atrium outlet opening area was selected to be equal to the total inlet openings areas for each storey. 3. Numerical solution procedures 3.1. CFD model The validated CFD model based on the use of the Reynolds Averaged NaviereStokes (RANS) modeling approach with the SSTkeu turbulence model and the Discrete Transfer Radiation Model (DTRM) presented in our previous studies, e.g. see Refs. [29e33] was utilized using the commercial CFD solver FLUENTÓ. The solution was obtained for representing the transfer of momentum and

2. Building description Previously the thermal environment in the atrium space of the Engineering Building of Concordia University, Montreal was studied by authors using a CFD model and the CFD predictions were validated against the experimental data available e.g., see

135

Fig. 1. A simple three-storied atrium building selected.

136

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146

Table 1 Dimensions and areas of the atrium building selected. Dimensions and areas Atrium height

12.00 m

Atrium width Atrium depth Exhaust chimney width Exhaust chimney height Height of inlets from ground and first floor Room height Room width Room depth Façade glazing area Ground floor air supply (net) area First floor air supply (net) area Second floor air supply (net) area Atrium outlet opening (net) area

5.00 m 6.00 m 2.00 m 6.00 m 1.1 m 4.00 m 6.00 m 6.00 m 60.00 m2 1.20 m2 1.08 m2 1.80 m2 4.08 m2

heat, buoyancy effects, turbulence, solar radiation, boundary conditions, and fluid properties. In the numerical modeling it has been assumed that (i) the flow is steady, turbulent and threedimensional, (ii) the flow is single phase, i.e., the effects of dust particles and water vapor are neglected, (iii) the flow at any inlet vent is uniform, (iv) the air properties are constant except for the density change with temperature that gives rise to the buoyancy forces, this being treated using the Boussinesq approach, and (v) external ambient conditions are steady. In dealing with the buoyancy forces in the momentum equations the Boussinesq approach was adopted, i.e., it was assumed that the fluid properties are constant except for the density change with temperature which gives rise to the buoyancy forces, these being dealt with a linear relation between the density change and the temperature change. In addition, the dissipation term in the energy equation was neglected due to the low velocities involved. The general form of the momentum, turbulent kinetic energy, turbulent energy dissipation, and energy (temperature for constant heat capacity) equations in the steady-state form can be expressed in the general form as follows:

  vðrui fÞ v vf Gf þ Sf ¼ vxi vxi xi

(8)

where variable (f) is (f) ¼ (u), (v), (w), (k), (ε), (h, T), Gf respectively, is diffusion coefficient of variable f and S4 represents the source terms including pressure terms, thermal source terms, etc., as appropriate for the variable (f) being solved. For pressure discretization, the body force weighted scheme was employed while the SIMPLE-algorithm was used for pressuree velocity coupling discretization. Second order upwind scheme was used to discretize the momentum and other equations in the numerical simulations. To get a converged solution, the equations for mass and momentum conservation were iteratively solved until the sum of the absolute normalized residuals for all the cells in flow domain became less than 104 while the energy equation was iterated until the residual fell below 106, the solution then being considered to be converged. Under-relaxation factors 0.3, 1, 2, 0.8, 0.8, 1, 0.9 for pressure, density, momentum, turbulence kinetic Table 2 Climatic parameters selected. Input parameters

Values

Shading Solar intensity Transmissivity Absorptivity Emissivity

Blinds open At 9:00 to 16:00 h on July 15, 2010 0.36 0.175 0.85

energy, turbulence dissipation rate, turbulent viscosity, energy respectively were used. In all the cases considered, the convergence criteria were met after about 10,000 iterations using a mesh size of approximately 820,000 cells. The mesh size was selected based on the results of the mesh independent tests. Simulations were performed on a window desktop PC with Intel Quad Core Processor EM64T Family 6 Model 15 Stepping 11 Genuine Intel w3.0 GHz each processor, 8 GB Ram 1333 Hz, Operating system is Windows XP Professional 64 bit SP2 and required approximately 24 h for the grid selected. 3.2. Radiation model In order to calculate the local solar heat gain in the building through the glass surfaces, the position of the sun in the sky must be known, and then the intensity of incident total solar radiation and the thermal radiation exchange between each building envelope element and the surroundings (sky, ground and neighboring buildings) must be calculated. This results in a source term that is added to the energy (temperature) equation. In the present work, solar calculator was used for calculation of the incident solar radiation. Total solar radiation incident on a sloped surface is:

It ¼ ID þ Ids þ Idg

(9)

where It is the total incident shortwave radiation, ID is the direct radiation, Ids the sky diffuse radiation, and Idg is the ground reflected radiation to the surface. Shortwave radiation incident upon a surface is considered to be converted to a thermal source, depending on the local absorption coefficient (a) of the material, which equals emissivity (e) for gray surfaces in equilibrium. To account for long-wave thermal radiation exchange among the surfaces within the building, radiation intensity transport equations (RTEs) are solved. FLUENT offers five radiation models: discrete transfer radiation model (DTRM); P-1 radiation model; Rosseland radiation model; surface-to-surface (S2S) radiation model; and discrete ordinates (DO) radiation model. From our previous studies it was noted that DTRM radiation model is more suitable for the present study. The main assumption followed in the DTRM model is that radiation leaving surface element in a specific range of solid angles can be approximated by a single ray. It uses a ray-tracing algorithm to integrate radiant intensity along each ray and is relatively simple model and increases accuracy by increasing number of rays while applies to a wide range of optical thicknesses. The solar calculator was used to find the sun’s location in the sky with the given inputs of time, date and the global location. Solar irradiation and outside conditions at 13:00 on July 15, 2010 are shown in Table 3. Table 3 Solar irradiation and outside conditions in Montreal at 13:00 h on July 15, 2010. Sun direction vector

Sunshine fraction Direct normal solar irradiation (at Earth’s surface) (W/m2) Diffuse solar irradiation e vertical surface (W/m2) Diffuse solar irradiation e horizontal surface (W/m2) Ground reflected solar irradiation e vertical surface (W/m2) Outside heat transfer coefficient (W/m2- C) Outside air temperature ( C)

x

y

z

0.54

0.84

0.06

1 (full shine) 868 141 116 91.05

7.4 29.7

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146

3.3. Boundary conditions

Total Solar Irradiation

1400

hc ¼ 4Vw þ 7:4

(10)

Outside Air Temperature

1200 Total Solar Irradiation (w/m2)

External climatic conditions affect architectural design of naturally ventilated buildings. The internal airflow pattern is the result of interaction between the indoor and outdoor environment. Especially in the case of natural ventilation, outdoor conditions strongly affect the indoor airflow pattern, and thus affect the thermal sensation of the occupants. In the present study the focus was to investigate the use of buoyancy-driven natural ventilation during day and night time, therefore numerical investigations were performed under fixed steady-state outdoor climatic conditions during daytime at hourly basis from 9:00 to 16:00 h and during night time at 1:00 h and 23:00 on 15 July 2010 which represents peak summer time for the location of region considered while wind velocity was assumed zero. The variation of external ambient temperature and total solar irradiation from 6:00 h to 20:00 h on July 15, 2010 in Montreal is shown respectively in Figs. 2 and 3. The natural ventilation of the building was considered based on the only buoyancy-driven natural ventilation induced by the heat gains from the solar radiation and other heat sources present in the building. The heat sources were assumed to be located in the center of each floor to match with the assumptions of the mathematical models developed by Holford and Hunt [35]. In practice airflow inside the connected spaces is mixed by conduction, convection and radiation heat transfer effects. In this work, conduction and radiation effects were only considered for the glazing facade walls while all the other walls were assumed to be adiabatic. All boundaries of the domain, except the façade glass surfaces, ventilation openings, and heat source were modeled as no-slip (uj ¼ 0) wall boundaries with zero heat flux. The mixed thermal boundary conditions were used for the façade glass surfaces. One of the most demanding aspects of heat transfer through the envelope is the evaluation of the convection heat transfer coefficient. There are many different correlations in the literature to determine the external heat transfer coefficient for buildings. Palyvos [36] summarized and outlined the different correlations found in literature. On the basis of thirty available linear correlations, he comes up with the following “average” correlation to calculate the heat transfer coefficient (hc) for windward surfaces:

137

1000 800 600 400 200 0 6:00

8:00

10:00

12:00

14:00

16:00

where Vw, is wind velocity. With zero wind velocity the value of external heat transfer coefficient equal to 7.4 W/m2 K was used. The optical properties of the glazing surfaces (semi transparent) used in our previous studies [29e34] were used with solar transmittance of 36% and absorptivity of 17.5%. The modeling of the glazing walls was simplified as a single glazing with effective thermal conductivity of 0.0626 W/m2 K with a total overall thickness 24 mm. The radiation exchange between the facade and the sky was also taken into account. The sky temperature was calculated using the Mills 4 1/4 ] where the emissivity of the sky [37] correlation, Tsky ¼ [εskyTout (εsky) for the daytime was calculated using the relation, εsky ¼ 0.727 þ 0.0060Tout. The heat sources were modeled as a noslip wall boundary (2  2 m) located on the center of each floor. A constant relative pressure of 0 Pa was imposed across the room inlets and the atrium outlet. 3.4. Mesh dependency test Three mesh densities with hexahedral cells were investigated: Mesh 1 (436,000 cells), Mesh 2 (832,000 cells, see Fig. 4) and Mesh 3 (1,209,000 cells). The grids used are highly non-uniform characterized by high-nodes density near solid walls and more cells are located where more velocity and temperature gradients are

30

Temperature (c)

25 20 15 10 5

8:00

10:00

12:00

14:00

16:00

18:00

20:00

Time (hr)

Fig. 2. Variation of external ambient temperature from 6:00 h to 20:00 h on July 15, 2010 in Montreal.

20:00

Fig. 3. Variation of total solar irradiation (W/m2) on hourly basis from 6:00 h to 20:00 h on July 15, 2010 in Montreal.

35

0 6:00

18:00

Tim e (hr)

Fig. 4. Mesh structure for CFD simulations (Mesh 2).

138

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146

expected, e.g., near walls, ventilation openings and the area potentially occupied by the thermal plumes in order to capture the variation in the airflows in these areas. The minimum cell size near each wall was selected, leading to yþ values 10 in order to apply SST-keu turbulence model. The simulation results shown in this sub-section are based on the conditions for achieving the same ventilation flow rate in each storey of the building. The volume flow rates at three floors using three mesh densities are shown in Table 4. It can be seen, there is very small difference less than 1% among the results obtained. Considering both accuracy and computational time, it was decided that Mesh 2 (812k hexahedral cells) is fine enough to accurately predict volume flow rate, airflow patterns and temperature distributions in the building. It took about 10,000 iterations for the cases simulated with a mesh size of 832,000 cells to meet the convergence criteria. 4. Results and discussion The proposed general ventilation concept involves fresh air being taken in from openings in the east, west and north-facing walls, passing through the occupant space, and flowing out from upper openings between the rooms and the atrium. The air is finally exhausted from the outlet on top of the building. It is assumed that the driving force is the temperature difference between the inside and outside temperatures of the building without considering the affects of external wind. Buoyancy-driven natural ventilation is more complex and difficult to model because of several parameters involved that are dependant and interdependent on the driving force of the ventilation. To cover all the parameters is beyond the scope of present paper. Here factors such as climate conditions, various geographical locations, orientation and internal loads were considered and will be discussed in detail to elucidate the performance of buoyancy-driven ventilation in the prototype residential building and other parameters will be investigated in future studies. 4.1. Validation of the CFD model The CFD model used in this work was validated in our recent studies, i.e., see Refs. [29e33] for the prediction of the airflow and temperature distributions in the atrium space of an Engineering building of Concordia University, Montreal, and an atrium building in Ottawa, Canada. The validation was carried out by comparing the CFD predictions against the experimental measurements. A close agreement was found between the CFD predictions and experimental measurements which successfully demonstrate the ability of the CFD model used to accurately predict three-dimensional buoyancy-driven displacement ventilation flows in multi-storey spaces connected to a common atrium. For the Concordia building, the average air temperatures along the height of the atrium were predicted at 16:00 h on August 1, 2007 and compared with the experimental data obtained by Mouriki [38]. For example, a comparison of the air temperatures predicted by the numerical model used and the measured temperature values at different locations in the atrium space of the Engineering building at Concordia University Montreal, Canada is shown in Table 5.

Table 4 Volume flow rates at three floors using three mesh densities. Floors

Volume flow rate (m3/s) Mesh 1

Mesh 2

Mesh 3

Ground floor First floor Second floor

0.42 0.42 0.40

0.42 0.42 0.40

0.42 0.42 0.40

The difference between the predictions and measurements has been expressed in terms of percentage error (%) [29]. A series of CFD simulations were run to investigate the various flow situations of buoyancy-driven ventilation in the building considered here. The results obtained are presented below. 4.2. Results for the daytime ventilation 4.2.1. Various flow situations considered Various flow situations for the buoyancy-driven ventilation investigated numerically are described here briefly and the ventilation airflow rates are calculated at supply inlets in all the cases considered: Case A: in this case the building was assumed to be located and oriented like the Engineering building at Concordia University, Montreal, Canada, i.e., 35 west of south, with the façade glazing surface facing southwest. In this case the effects of variation in ambient temperature and incidence of total solar irradiation on the development of buoyancy-driven volume flow rates in the building were explored without considering the internal heat sources. Case B: in this case the effects of variation in internal heat sources on the development of buoyancy-driven volume flow rates in the building were studied at a particular time of the day. Case C: this case explores the effect of the geographical location on the ventilation airflow rate in the building. In this case the building was assumed to be located in various places in Canada: St. John’s, Montreal, and Calgary with the orientation of the building, i.e., 35 west of south, with the façade glazing surface facing southwest. Case D: this case presents the effect of the variation in orientation of the building on the ventilation volume flow rates. In this case the building was assumed to be oriented with the glazing surface facing south, south east and southwest, located in Montreal, Canada. Simulations were run to investigate the building performance for various flow situations described above involving the buoyancydriven ventilation in the building. The results obtained are presented below. 4.2.1.1. Results for daytime buoyancy-driven ventilation in the building 4.2.1.1.1. Case A: outdoor climate conditions. The volume airflow rates in the rooms and atrium space calculated at supply inlets are evaluated using different ambient temperatures and solar irradiation at 9:00 to 16:00 h on July 15, 2010 which represents the peak summer day in Montreal, Canada. The simulation results, which are shown in Table 6a and b reveal that different outdoor temperatures and solar irradiation have significant influences on indoor temperature distributions and airflow rates. In addition, the results indicate that when the outdoor temperature is less than 30  C, the buoyancy-driven natural ventilation technique can be utilized for the occupied spaces to reach the general thermal comfort level. However, other ventilation techniques are needed when the outdoor temperature is greater than 30  C to reach the same comfort level. 4.2.1.1.2. Case B: indoor thermal load conditions. Here the volume airflow rates in the rooms and atrium space are evaluated using different heat-loading conditions. In this part of the simulation, an ambient temperature of 28  C at 11:00 h on July 15, 2010 in Montreal is used. Heat sources of 100, 200, 300, 400 and 500 W/m2 were assumed which cover an area of 2  2 m in the center of each floor of the building considered. All other walls except the façade

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146

139

Table 5 Comparison of the numerical predictions and experimental measurements of indoor air temperatures using SST-keu turbulence model. Y (m)

Temperature T ( C) x ¼ 5.96, z ¼ 7 m

2.1 6.16 10.25

x ¼ 5.78, z ¼ 1.05 m

x ¼ 8.81, z ¼ 4.447 m

Measured

Predicted

% Error

Measured

Predicted

% Error

Measured

Predicted

% Error

24 24.5 26.2

22.87 24.55 26.52

4.7 0.2 1.1

22.3 25.1 26.1

22.37 24.47 26.20

0.3 2.5 0.4

23 24.6 26.3

22.71 24.72 26.55

0.5 0.4 0.3

glazing wall were assumed adiabatic. The simulation results for the thermal load conditions are presented in Table 7a, b. The results show that the change in airflow rates in the rooms and the atrium is less than 0.05 m3/s which in not as significant as the change in airflow rates caused by outdoor climate conditions discussed earlier. The comparative results shown in Table 6 reveal that in the critical case, namely a high outdoor temperature and high internal load, extra ventilation strategies are needed for the building occupied space to reach the required comfort level. Fig. 5(aed) shows the contours of pressure, velocity, turbulence intensity, and temperature distributions in the middle plane parallel to the facade surface along the height of the building respectively with a heat source of 200 W/m2 at each occupied area of the rooms and atrium floor at 13:00 h on July 15, 2010. Fig. 5a shows stack pressure developed which is the pressure difference between the interior and exterior of the building created naturally by density differences caused by warm air generated inside the building due to solar heat gain and heat sources present in the building. This pressure difference drives ventilation flow through the building by means of thermal stack effects. Fig. 5b shows velocity contours of buoyancydriven ventilation airflow and indicates the influence of gravitational body forces on the velocity fields within the rooms and the atrium space. It was seen that in the presence of solar radiation, free convection effects are shown by the formation of air stream moving upward around the hot surfaces. Strong stack effect creates higher inflow on the lower floors than the upper floors. Fig. 5c presents turbulence intensity contours in the middle plane parallel to the facade surface along the height of the building. The instantaneous velocity vector (u) can be decomposed into mean and turbulent components in the X, Y and Z directions:

ui ¼ Ui þ u0i

(11)

where Ui is the mean velocities and ui0 is their turbulent counterparts, with index i representing X, Y and Z. The r.m.s. of each veis defined as the standard deviation of ui locity component, ur.m.s. i and the normalized r.m.s. velocity or turbulence intensity is defined

as Ii ¼ ur.m.s. /Ui. The results show that turbulence intensity values i are nearly uniform along the vents with a tendency of higher values near the outflow than inflow and in the regions where there is more mixing and circulation of the airflow within the building. Fig. 5d shows temperature distribution within the building. Higher air temperature stratification was observed near the ceilings of the rooms and near the outlet in the chimney walls. Lower temperature stratification near the floors was due to the strong convection. 4.2.1.1.3. Case C: geographical locations. To examine the effect of the geographical location on the ventilation airflows in the building, the building was assumed to be located in different cities in Canada: St. John’s, Montreal, Winnipeg, and Calgary with the same orientation. The numerical results obtained for the volume airflow rates of the buoyancy-driven ventilation are shown in Table 8. 4.2.1.1.4. Case D: orientation of the building. To study the effect of the orientation of the building on the volume flow rates of airflows, it was assumed to be oriented with glazing surface facing south, south east and southwest, located in Montreal, Canada. The numerical results obtained for the volume airflow rates of the buoyancy-driven ventilation in the building are given in Table 9. It is seen that the building orientation with glazing surface facing southward has better performance. From the simulations results obtained, it is evident that for buoyancy-driven ventilation, the outdoor airflows into the indoor space and mixes with indoor air. The performance of buoyancydriven ventilation depends on the ventilation period and the temperature difference between indoor and outdoor spaces. The outdoor temperatures have a significant effect on the indoor thermal performance. The temperature distribution shown in Fig. 5d shows that inside temperatures are 0.5e10 higher than the outside temperatures. It is noted that to meet the basic comfort requirement defined by ASHRAE Standard 55-2004 (namely, 18e24  C), the ambient air temperature must be 10  C lower than the air temperature within the building. Moreover, in hot and humid weather, more effort has to be made to reach the same indoor comfort level under the period of limited temperature difference between indoor and outdoor spaces.

Table 6 (a) Buoyancy-driven volume airflow rate in the left-hand side rooms and atrium space of the building located in Montreal at 9:00 to 16:00 h on July 15, 2010. (b) Buoyancydriven volume airflow rate in the left-hand side rooms and atrium space of the building located in Montreal at 9:00 to 16:00 h on July 15, 2010. Volume flow rate (m3/s) (9:00)

Volume flow rate (m3/s) (10:00)

Volume flow rate (m3/s) (11:00)

Volume flow rate (m3/s) (12:00)

Volume flow rate (m3/s) (13:00)

Volume flow rate (m3/s) (14:00)

Volume flow rate (m3/s) (15:00)

Volume flow rate (m3/s) (16:00)

(a) Ground floor First floor Second floor Atrium

0.42 0.42 0.42 0.80

0.41 0.41 0.42 0.78

0.40 0.39 0.36 0.75

0.36 0.36 0.33 0.69

0.31 0.31 0.34 0.56

0.25 0.26 0.28 0.45

0.25 0.26 0.28 0.45

0.24 0.25 0.26 0.42

(b) Ground floor First floor Second floor

0.43 0.43 0.42

0.42 0.42 0.39

0.40 0.39 0.36

0.36 0.36 0.35

0.31 0.30 0.32

0.26 0.27 0.32

0.27 0.28 0.31

0.26 0.27 0.30

Left-hand side rooms

140

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146

Table 7 (a) Buoyancy-driven volume airflow rate with variation in internal heating load in the left-hand side rooms and atrium space of the building located in Montreal at 11:00 h on July 15, 2010. (b) Buoyancy-driven volume airflow rate with variation in internal heating load in the right-hand side rooms of the building located in Montreal at 11:00 h on July 15, 2010. Left-hand side rooms (a) Ground floor First floor Second floor Atrium (b) Ground floor First floor Second floor

Volume flow rate (m3/s) (100 W/m2)

Volume flow rate (m3/s) (200 W/m2)

Volume flow rate (m3/s) (300 W/m2)

Volume flow rate (m3/s) (400 W/m2)

Volume flow rate (m3/s) (500 W/m2)

0.43 0.42 0.40 0.79

0.44 0.43 0.42 0.83

0.46 0.46 0.45 0.84

0.47 0.47 0.46 0.88

0.49 0.49 0.48 0.90

0.43 0.43 0.41

0.44 0.44 0.43

0.46 0.46 0.44

0.47 0.47 0.46

0.49 0.48 0.47

4.3. Evaluation of the thermal comfort conditions in the building at 13:00 on 15 July 2010 The thermal comfort conditions in the building shown in Fig. 1 were evaluated using the thermal comfort indices. The values of the air temperatures and velocities (V) in the area of interest were determined from CFD simulations. The values of the PMVNV and PPDNV (%) were calculated in the occupied area along the horizontal plane at nine points (x ¼ 1, 3, 5 and z ¼ 1, 3, 5) for seated activity at 0.6 m above each floor using the JAVA applet ISO 7730 computer program [26]. The values of activity, metabolic rate (W/m2), clothing (clo) and relative humidity (%) were selected 0, 60, 0.5 and 54, respectively. The expectancy factor was estimated to be 0.8 in

the calculations of PMVNV and PPDNV and the results obtained are shown in Table 10aed. From the results given in these tables, it will be seen that for selected points considered in the occupied areas, the values of PMVNV are in the range 0.18 to 0.55 (corresponding to neutral comfortable) on the ground, first and atrium floors and in the range 0.73 to 0.80 (slightly cool, acceptable) on the second floor. The PPDNV (%) values indicate that only 5e16% people will not be satisfied with the thermal conditions on the ground, first and atrium floors but that on the second floor 16e18% of the people will be dissatisfied with the thermal conditions. Overall, it will be seen the thermal conditions existing in the modified design of the building due to buoyancy-driven ventilation are mostly in comfortable zone.

Fig. 5. (aed) Pressure (a), velocity (b), turbulence intensity (c) and temperature (d) contours in the middle plane parallel to the façade surface along the height of the building.

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146 Table 8 Ventilation volume flow rates at each airflow inlet in various locations at 13:00 h on July 15, 2010. Geographical locations

St. John’s Montreal Calgary

Volume flow rate (m3/s) at each airflow inlet LHS rooms

RHS rooms

Atrium

Ground floor

First floor

Second floor

Ground floor

First floor

Second floor

0.46 0.46 0.45

0.47 0.45 0.45

0.46 0.47 0.46

0.46 0.46 0.45

0.46 0.47 0.46

0.47 0.46 0.47

0.85 0.87 0.85

Table 9 Ventilation volume flow rates at each airflow inlet with different orientation of the building. Building orientation with façade glazing surface facing 45 Southwest South 45 Southeast

Volume flow rate (m3/s) at each airflow inlet LHS rooms

RHS rooms

Atrium

Ground floor

First floor

Second floor

Ground floor

First floor

Second floor

0.44 0.47 0.45

0.43 0.46 0.45

0.48 0.49 0.46

0.44 0.48 0.45

0.44 0.46 0.46

0.51 0.51 0.47

0.81 0.86 0.85

4.4. Discomfort due to draft For the assessment of draft discomfort, the thermal comfort empirical Eq. (11) to calculate the percentage dissatisfied (PD) given by Fanger and Toftum [27] and adopted in ISO 7730 [28] was used.

PDðpercent dissatisfiedÞ ¼ ð34  Ta ÞðV  0:5Þ0:62  ð0:37V  Tu þ 3:14Þ

(12)

This equation requires the values of three parameters; velocity (V), air temperature (Ta) and turbulence intensity (Tu) to be known. These parameters were evaluated at a height of 0.6 m for seated activity. Fig. 6aec shows the contours of the contours of temperature, velocity and turbulence intensity respectively obtained through CFD simulations over a horizontal plane at 0.6 m above each floor in the occupied areas of the building. Using the values of the thermal comfort parameters shown in Fig. 6aec in Eq. (3), the contours of PD factor were calculated over the region

141

considered and displayed in Fig. 6d. From Fig. 4d it is seen that on the second floor near the airflow inlets, the PD values are higher due to both low temperatures and high velocities. However, in the whole region of the occupied area of second floor, the PD values lie in the range of 2e10%. The calculated values of the PD (%) in the occupied area of the first floor and atrium space are in the range of 2e15% which indicates that due to draft the thermal conditions developed by the solar-assisted buoyancy-driven airflows are in comfortable zone and only a small percentage of less than 15% of the occupants is expected to be slightly uncomfortable due to draft. 4.5. Results for buoyancy-driven night ventilation and thermal comfort evaluation in the building 4.5.1. Night ventilation flow situations simulated The night ventilation of buildings can be incorporated as a part of the passive ventilation strategy, with the aim being to purge the building of warm air that has accumulated during the day and, thereby, to cool the building. Several studies have shown that night ventilation is effective in improving indoor thermal comfort conditions at night and reducing the cooling demand during the daytime. The purpose of present work was to gain insight into this behavior of a simple three-storey atrium building using CFD modeling techniques. Various cases involving buoyancy-driven night ventilation were investigated numerically, these cases being briefly described below: Case A: in this case the development of buoyancy-driven night ventilation induced by the heat sources present on each floor of the building was considered. Case B: in this case the night ventilation induced by hot water at 80  C flowing in the chimney walls and in central plate, both being 3 m high in the chimney was considered. Case C: this case presents the night ventilation induced by both heat sources on each floor and plus hot water at 80  C of the chimney walls and central plate. Case D: this case presents the night ventilation induced by both heat sources on each floor and plus hot water at 60  C of the chimney walls and central plate.

Table 10 (aed) calculations of PMVNV and PPDNV for seated activity at 0.6 m above each floor in the occupied area (a) ground floor (b) first floor (c) second floor and (d) atrium floor. Z, m

X¼1m

X¼3m

T ( C)

V (m/s)

PMVNV

PPDNV

(a) 1 3 5

29.56 29.75 29.3

0.09 0.42 0.13

0.45 0.27 0.48

(b) 1 3 5

29.4 29.56 29.18

0.14 0.18 0.11

(c) 1 3 5

27.5 27.2 27.5

(d) 1 3 5

31.14 30.4 29.32

X¼5m

T ( C)

V m/s

PMVNV

PPDNV

T ( C)

V m/s

PMVNV

PPDNV

8.4 6.7 9.7

29.2 27.5 29.2

0.16 0.66 0.16

0.44 0.60 0.44

9.2 16.2 9.2

28.7 28.5 28.9

0.08 0.33 0.11

0.56 0.54 0.53

11 10.8 10.6

0.46 0.39 0.50

9.5 8.1 10.3

28.7 28.8 28.9

0.14 0.31 0.08

0.56 0.48 0.54

11 10.2 10.8

28.3 28.6 28.6

0.09 0.28 0.11

0.60 0.52 0.58

12.7 10.4 11.2

0.08 0.10 0.06

0.73 0.77 0.73

15.9 16.4 15.9

27.2 27.3 27.2

0.10 0.13 0.09

0.77 0.76 0.77

16.4 16.3 16.4

27.1 26.8 27.1

0.09 0.02 0.05

0.79 0.83 0.79

18.4 19.3 18.4

0.08 0.32 0.36

0.23 0.18 0.38

6.3 5.4 8.0

30.3 28.16 28.8

0.32 0.66 0.27

0.18 0.58 0.50

5.4 11.2 10.2

30.4 30.3 29.6

0.18 0.18 0.27

0.25 0.24 0.34

6.3 6.2 7.8

142

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146

Fig. 6. Temperature (a) velocity (b) turbulence intensity (c) and PD (d) contours along the horizontal planes at the height of 0.6 m from each floor in the occupied areas of the building considered.

Case E: this case presents the night ventilation induced by both heat sources on each floor and plus hot water at 40  C of the chimney walls and central plate. Case F: in this case the night ventilation induced by heat sources on each floor and by hot water at 80  C flowing in the chimney walls, i.e., without the central heated plate is investigated. Simulations were run for all of the cases considered. The volume flow rates calculated at each inlet for these various cases considered at 2: 00 h on July 15, 2010 with ambient temperature of 22  C are given in Table 11. The total heat transfer rates calculated for each case are given in Table 12. A comparison of the volume flow rates (m3/s) in the right-hand side rooms for the different cases considered is shown in Fig. 7 From Fig. 7, it can be seen that ventilation airflow rate is higher for the Case C where night ventilation is induced by the heat sources and by hot water heating at 80  C of the chimney walls and of a central plate. Contours of pressure, velocity, turbulence intensity, and temperature over the middle plane parallel to the façade glazing surface of the building for buoyancy-driven night ventilation for Case C are shown in Fig. 8aed respectively. From these results it will be seen that the chimney heating contributes considerably to the ventilation flow rate. This indicates

that if chimney heating is to be used a larger heater surface should be the better option. Case G: indoor thermal load conditions The volume airflow rates in the rooms and atrium space during buoyancy-driven night ventilation are evaluated using different

Table 11 Volume flow rates at each inlet for various cases of buoyancy-driven night ventilation at 02:00 h on July 15, 2010 with ambient temperature of 22  C. Different flow situations of buoyancy-driven night ventilation considered

Volume flow rate (m3/s) at each airflow inlet

Ground floor

First floor

Second floor

Ground floor

First floor

Second floor

Case Case Case Case Case Case

0.25 0.09 0.28 0.27 0.26 0.26

0.25 0.12 0.29 0.28 0.26 0.26

0.30 0.19 0.35 0.33 0.33 0.36

0.25 0.09 0.28 0.27 0.26 0.28

0.25 0.12 0.29 0.28 0.26 0.28

0.31 0.13 0.37 0.39 0.33 0.36

A B C D E F

LHS rooms

RHS rooms

Atrium

0.45 0.17 0.48 0.47 0.43 0.47

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146

143

Table 12 Total heat transfer rate for various cases of buoyancy-driven night ventilation. Different flow situations considered

Total heat transfer rate (W) Ground floor heat source

First floor heat source

Second floor heat source

Atrium heat source

Chimney walls hot water

Chimney plate hot water

Total

Case Case Case Case Case Case

1640 e 1640 1640 1640 1640

1640 e 1640 1640 1640 1640

1640 e 1640 1640 1640 1640

520 e 520 520 520 520

e 2683 4354 2127 1250 7997

e 1170 2153 941 725 e

5440 3862 11,947 8508 7416 13,437

A B C D E F

0.4

Volume flow rate (m3/s)

0.35

heat-loading conditions. In this part of the simulation, an ambient temperature of 26.6  C at 23:00 h on July 15, 2010 in Montreal is used. The heat sources of 100, 200, 300, 400 and 500 W/m2 were assumed which cover an area of 2  2 m in the center of each floor of the building considered. All other walls except the facade glazing wall, chimney walls and chimney central plate were assumed adiabatic. The simulation results of the thermal load conditions are presented in Table 13a, b. The results show that the change in airflow rates in the rooms and the atrium is less than 0.07 m3/s which in not as significant as the change in airflow rate caused by outdoor climate conditions discussed earlier. The comparative results shown in Table 13a, b reveal that in the critical case, namely a high outdoor temperature and high internal load, extra ventilation strategies are needed for the building occupied space to reach the required comfort level.

Ground floor First floor Second floor

0.3 0.25 0.2 0.15 0.1 0.05 0 Case-A

Case-B

Case-C

Case-D 3

Case-E

Case-F

Fig. 7. Comparison of the volume flow rates (m /s) in the right-hand side rooms for the different cases of night ventilation considered.

4.5.2. Thermal comfort evaluation in the building (case C) The thermal comfort conditions in the building were evaluated using the thermal comfort indices and discomfort due to draft. The

Fig. 8. (aed) Contours of pressure, velocity, turbulence intensity and temperature in middle plane along the height of the building during night ventilation for the Case C.

144

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146

Table 13 (a) Buoyancy-driven volume airflow rate with variation in internal heating load in the left-hand side rooms and atrium space of the building located in Montreal at 23:00 h on July 15, 2010. (b) Buoyancy-driven volume airflow rate with variation in internal heating load in the right-hand side rooms of the building located in Montreal at 23:00 h on July 15, 2010. Left-hand side rooms

Volume flow rate (m3/s) (100 W/m2)

Volume flow rate (m3/s) (200 W/m2)

Volume flow rate (m3/s) (300 W/m2)

Volume flow rate (m3/s) (400 W/m2)

Volume flow rate (m3/s) (500 W/m2)

(a) Ground floor First floor Second floor Atrium

0.22 0.24 0.28 0.37

0.28 0.29 0.36 0.48

0.31 0.32 0.38 0.55

0.34 0.36 0.39 0.61

0.37 0.39 0.43 0.66

(b) Ground floor First floor Second floor

0.21 0.23 0.31

0.28 0.29 0.36

0.31 0.32 0.38

0.34 0.36 0.40

0.36 0.38 0.42

PMVNV and PPDNV indices were calculated using six basic variables: activity, clothing, air temperature, air velocity, mean radiant temperature (MRT), and relative humidity (%). The mean values of the air temperatures (T) and velocities (V) in the area of interest were determined from the CFD simulations. The values of the PMV and PPD (%) were calculated for seated activity in the occupied area along the horizontal plane at nine points (x ¼ 1, 3, 5 m and z ¼ 1, 3, 5 m) at a height of 0.6 m above each floor using the JAVA applet ISO 7730. The values of activity, metabolic rate (W/m2), clothing (clo) and relative humidity (%) were selected 0, 60, 0.9 and 70 respectively in the calculations of PMVNV and PPDNV and an expectancy factor was estimated to be 0.80. The results obtained are given in Table 14aed. From the results given in these tables, it will be seen that that for the selected points in the occupied area of each floor, the values of PMVNV are in the range 0.35 to 0.50 (corresponding to neutral comfortable). The PPDNV (%) values indicate that only 7e9% people will not be satisfied with the thermal conditions on each floor. Overall, it will be seen the thermal conditions existing in the building due to buoyancy-driven night ventilation are in comfortable zone. 4.5.3. Discomfort due to draft in the building for night time ventilation for case C For the assessment of the draft discomfort the thermal comfort empirical was used to calculate the percentage dissatisfied (PD) as

given by Fanger and Toftum (1970) and adopted in ISO 7730 [28]. This equation requires that the values of three parameters; air velocity (V), air temperature (Ta) and air turbulence intensity (Tu) be known. These parameters were evaluated at a height of 0.6 m for seated activity above each floor. Fig. 7aec shows the contours of these three required parameters over a horizontal plane at a height of 0.6 m above each floor in the building. Using the values of the variables shown in Fig. 6aec, the contours of the thermal comfort parameter were obtained and are shown in Fig. 6d. From the Fig. 6d, it will be seen that the PD values for most of the occupied region are in the range 4e8%. However, in the regions near the inlets on the ground, first and atrium floors, the PD values are in the range 16e30%. The calculated values of the PD indicate that most of the occupied area of the rooms and atrium space ventilated by buoyancy-driven night ventilation airflow is in comfortable zone. Only a small percentage of the occupants can be expected to be slightly uncomfortable due to draft. From the present work, it is concluded that the thermal conditions in the occupied areas of the building developed as a result of the use of buoyancy-driven ventilation for the particular values of the design parameters selected are mostly in the comfortable zone. Finally, it was demonstrated that the proposed methodology led to reliable thermal comfort predictions, while the effect of various design variables on the performance of the building was easily recognized.

Table 14 (aed) Calculations of PMVNV and PPDNV for seated activity at 0.6 m above each floor in the occupied area (a) ground floor (b) first floor (c) second floor and (d) atrium floor. Z, m

X¼1m

X¼3m

X¼5m

T ( C)

V (m/s)

PMVNV

PPDNV

T ( C)

V, m/s

PMVNV

PPDNV

T, ( C)

V, m/s

PMVNV

PPDNV

(a) 1 3 5

22.4 21.4 22.4

21 21.5 21

0.03 0.31 0.08

0.32 0.50 0.32

7.7 15.2 7.7

22.4 22.0 21.2

21 21 21.5

0.05 0.07 0.05

0.32 0.36 0.41

7.7 8.1 8.2

22.7 22.1 21.7

21 21 21.5

(b) 1 3 5

22.6 20.5 22.5

21 22.5 21

0.07 0.47 0.09

0.28 0.40 0.30

6.7 8.2 7.4

22.1 22.2 22.2

21 21 21

0.02 0.07 0.05

0.35 0.34 0.34

7.6 7.4 7.4

22.1 22.2 22.1

21 21 21

(c) 1 3 5

21.5 21.5 21.5

21.5 21.5 21.5

0.03 0.03 0.01

0.37 0.37 0.37

7.9 7.9 7.9

21.5 21.5 21.4

21.5 21.5 21.5

0.02 0.02 0.04

0.37 0.37 0.37

7.9 7.9 7.9

21.5 21.5 21.4

21.5 21.5 21.5

(d) 1 3 5

21.2 21.2 21.3

22 22 22

0.10 0.13 0.09

0.36 0.36 0.35

7.7 7.7 7.4

20.8 20.5 21.0

22 22.5 22.1

0.43 0.92 0.14

0.40 0.30 0.39

8.4 7.4 8.2

21.1 21.3 21.5

22 22 22

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146

5.

Fig. 9. Temperature distribution profiles in the center of the atrium floor along the height of the atrium from 0 to 2 m at 11:00 a.m. and 11:00 p.m.

4.5.4. Vertical air temperature difference Another cause of thermal discomfort is the vertical air temperature differences that occupants are experiencing. For all the cases considered, the temperature difference between the temperatures at the ankles, chest and head for a standing person at the height of 0.1, 1.1 and 1.5 m were calculated. Fig. 9 shows temperature distribution profiles in the center of the atrium floor along the height of the atrium from 0 to 2 m at 11:00 a.m. and 11:00 p.m. It was found that temperature difference is less than 2  C, which corresponds to a PD of less than 5% [28]. This demonstrates that vertical air temperature differences are moderate and should not cause major comfort issues for the occupants.

6.

7.

8. 5. Conclusions This research work has established that a validated CFD model can predict the pressure differences produced by warm air that has accumulated in an atrium building that drive ventilation airflows, in which warm air from the stories discharges into the atrium as turbulent thermal plumes. The model predicts the depth and temperature of the layers of warm air in each storey and in the atrium space. Various cases of night ventilation based on the buoyancy-driven airflow induced by heat sources in the occupied areas of the building and hot water flowing along the chimney walls were explored. Furthermore the thermal comfort conditions in the building were evaluated in terms of well-known thermal comfort indices. The following main conclusions were drawn from this study. 1. The design curves [35] are very useful in establishing the sizes of inlets and outlets in order to have equal ventilation flow rates in each storey of the atrium building. 2. The temperature change and volume airflow rate change caused by the internal thermal load is not as significant as that caused by outdoor climate conditions. Simulation results show that the external ambient temperature has a greater influence on the buoyancy-driven airflow rate and on the temperature distribution inside the building. The variation in internal thermal loads have little influence (1  C or 0.05 m3/s) on the thermal environment of the building. 3. In hot and humid weather, a building would require a significant height in order to induce a sufficient pressure gradient caused by temperature difference for efficient buoyancy-driven ventilation. 4. Thermal comfort conditions for daytime produced as a result of solar-assisted buoyancy-driven ventilation in the building were evaluated in details in terms of the thermal comfort indices

9.

145

(PMVNV, PPDNV and PD) referring to both skin-bulk flow-net heat balance and air draughts for seated activity. It was concluded that the thermal conditions developed in the building as a result of the daytime buoyancy-driven ventilation are neutral comfortable on the ground floor, the first floor and the atrium floor while on the second floor they are slightly cool acceptable. Thermal conditions due to draft and temperature variations developed by the solar-assisted buoyancy-driven airflows are in comfort zone and only a small percentage of less than 15% of the occupants is expected to be slightly uncomfortable due to draft. Thermal comfort conditions for night ventilation induced as a result of local heat sources and the circulation of hot water along chimney walls and a central plate were evaluated in terms of thermal comfort indices (PMVNV, PPDNV and PD) referring to both skin-bulk flow-net heat balance and air draughts for seated activity. It was found that the thermal conditions developed in the building as a result of the buoyancy-driven night ventilation are neutral comfortable on all floors. The thermal conditions developed due to draft by the buoyancy-driven night ventilation airflows are in comfortable zone and only a small percentage in the atrium space of less than 25% of the occupants is expected to be slightly uncomfortable due to draft. The results of the study showed that the buoyancy-driven ventilation provide acceptable comfort conditions inside the atrium building which can be maintained by exhausting relief air from the building through the atrium and chimney. From the numerical results obtained, it was found that the hot water heated by solar collectors during daytime can be effectively utilized for night ventilation as a part of the passive ventilation strategy. From the results it was found that temperature T ( C) rate. This indicates that if chimney heating is to be used a larger heating surface than considered here should be used. The results of this study can be considered at the initial design stage of an energy-efficient atrium building in order to obtain a comfortable indoor thermal environment. Based on this study, future research will include wind speed and other climatic factors.

Overall, the proposed methodology provides a useful procedure for building designers to quantify the buoyancy-driven natural ventilation at the initial design stage of an energy-efficient atrium building. It should be mentioned that the CFD modeling approach introduced is general and it can meet thermal comfort estimation produced by either turbulence modeling or experimental measurements. Acknowledgements This work was supported by the Natural Sciences and Engineering Research Council Canada (NSERC). References [1] M. Santamouris, K. Pavlou, A. Synnefa, K. Niachou, D. Kolokotsa, Recent progress on passive cooling techniques and advanced technological developments to improve survivability level in low-income households, Energy and Buildings 39 (2007) 859e866. [2] J. Vander Maas, C.A. Roulet, Night Time Ventilation by Stack Effect, ASHRAE Technical Data Bulletin, vol. 7 (1), Ventilation and Infiltration, NY, January 1991, pp. 32e40. [3] D. Finn, D. Connolly, P. Kenny, Sensitivity analysis of a maritime located night ventilated library building, Solar Energy 81 (2007) 697e710. [4] V. Geros, M. Santamouris, A. Tsangrasoulis, G. Guarracino, Experimental evaluation of night ventilation phenomena, Energy and Buildings 29 (1999) 141e154. [5] G. Guohui, Simulation of buoyancy induced flow in open cavities for natural ventilation, Energy and Buildings 38 (2006) 410e420.

146

S. Hussain, P.H. Oosthuizen / Applied Thermal Engineering 57 (2013) 133e146

[6] M.J. Cook, Y. Ji, G.R. Hunt, CFD modeling of natural ventilation: combined wind and buoyancy forces, International Journal of Ventilation 1 (2003) 169e180. [7] Y. Ji, M.J. Cook, V. Hanby, CFD modelling of natural displacement ventilation in an enclosure connected to an atrium, Building and Environment 42 (2007) 1158e1172. [8] Y. Jiang, C. Allocca, Q. Chen, Validation of CFD simulations for natural ventilation, International Journal of Ventilation 2 (4) (2004) 359e370. [9] T. Josef, H. Vitaly, T. Meir, Airflow and heat flux through the vertical opening of buoyancy-induced naturally ventilated enclosures, Energy and Buildings 40 (2008) 637e646. [10] H.L. Kwang, K.S. Richard, Enhancement of natural of natural ventilation in buildings using a thermal chimney, Energy and Buildings 41 (2009) 615e621. [11] P. Liu, H. Lin, J. Chou, Evaluation of buoyancy-driven ventilation in atrium buildings using computational fluid dynamics and reduced-scale air model, Building and Environment 1 (10) (2009). [12] L. Yang, Y. Li, Cooling load reduction by using thermal mass and night ventilation, Energy and Buildings 38 (2005) 959e963. [13] V. Geros, M. Santamouris, S. Karatasou, A. Tsangrassoulis, N. Papanikolaou, On the cooling potential of night ventilation techniques in the urban environment, Energy and Buildings 37 (2005) 243e257. [14] M. Santamouris, A. Sfakianaki, K. Pavlou, On the efficiency of night ventilation techniques applied to residential buildings, Energy and Buildings 42 (2010) 1309e1313. [15] J. Pfafferott, S. Herkel, M. Jaschke, Design of passive cooling by night ventilation: evaluation of a parametric model and building simulation with measurements, Energy and Buildings 35 (2003) 1129e1143. [16] R. Letan, V. Dubovsky, G. Ziskind, Passive ventilation and heating by natural convection in a multi-storey building, Building and Environment 38 (2003) 197e208. [17] N.K. Bansal, R. Mathur, Solar chimney for enhanced stack ventilation, Building and Environment 28 (1993) 373e377. [18] B. Givoni, Effectiveness of mass and night ventilation in lowering the indoor daytime temperature, Energy and Buildings 28 (1998) 25e32. [19] C. Afonso, A. Oliveira, Solar chimneys: simulations and experiment, Energy and Buildings 32 (2000) 71e90. [20] A. Atif, D. Claridge, Atrium buildings, thermal performance and climatic factors, ASHRAE Transactions 101 (1) (1995) 454e460. [21] M. Kavgic, D. Mumovic, Z. Stevanovic, A. Young, Analysis of thermal comfort and indoor air quality in a mechanically ventilated theatre, Energy and Buildings 40 (2008) 1334e1343. [22] R. Li Adrian, Thermal comfort and environmental modeling in atrium buildings, in: PLEA2006, 23rd conference on passive and low energy architecture, Geneva, Switzerland (6e8 September 2006). [23] K.W.D. Cheong, E. Djunaedy, Y.L. Chua, K.W. Tham, S.C. Sekhar, N.H. Wong, M.B. Ullah, Thermal comfort of an air-conditioned lecture theatre in the tropics, Building and Environment 38 (2003) 63e73. [24] ASHRAE, ASHRAE Handbook Fundamentals, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, USA, 1997. [25] Available from: www.healthyheating.com. [26] Available from: http://www.eat.lth.se/fileadmin/eat/Termisk_miljoe/PMVPPD.html. [27] P.O. Fanger, J. Toftum, Extension of the PMV model to non-air-conditioned buildings in warm climates, Energy and Buildings 34 (6) (2002) 533e536. [28] ISO 7730. Ref no ISO 7730, in: Moderate Thermal Environments: Determination of the PMV and PPD Indices and Specifications of the Conditions for Thermal Comfort, second ed., International Standards Organization, Geneva, 1994. [29] S. Hussain, P.H. Oosthuizen, An evaluation of turbulence models for the numerical study of flow and temperature distribution in atria, in: 18th annual conference of the CFD Society of Canada, London, Ontario (17e19 May 2010). [30] S. Hussain, P.H. Oosthuizen, Validation of numerical study of flow and temperature distribution in an atrium space, in: 19th annual conference of the CFD Society of Canada Montreal, Canada (27e29 April 2011).

[31] S. Hussain, P.H. Oosthuizen, Numerical study of an atrium integrated with hybrid ventilation system, in: 23rd Canadian congress of applied mechanics, Vancouver, BC, Canada (5e9 June 2011). [32] S. Hussain, P.H. Oosthuizen, Validation of numerical modeling of conditions in an atrium space with a hybrid ventilation system, Building and Environment 52 (2012) 152e161. [33] S. Hussain, P.H. Oosthuizen, A. Kalendar, Evaluation of various turbulence models for the prediction of the airflow and temperature distributions in atria, Energy and Buildings 48 (2011) 18e28. [34] Shafqat Hussain, Patrick Oosthuizen, Numerical study of the effect of atrium design changes on the flow rate and temperature distributions in a simple atrium building and thermal comfort evaluation, Applied Thermal Engineering 40 (2012) 358e372. [35] J.M. Holford, G.R. Hunt, Fundamental atrium design for natural ventilation, Building and Environment 38 (2003) 409e426. [36] J.A. Palyvos, A survey of wind convection coefficient correlations for building envelope energy systems modeling, Applied and Thermal Engineering 28 (2008) 801e810. [37] A.F. Mills (Ed.), Heat Transfer, second ed,, Prentice Hall, New Jersey, 1999, pp. 570e572. [38] E. Mouriki, Solar Assisted Hybrid Ventilation in an Institutional Building, (M.Sc. thesis), Department of Building, Civil and Environmental Engineering, Concordia University, Montreal, Canada, 2009.

Glossary e: expectancy factor fcl: garment insulation factor g: gravitational acceleration (m s2) hc: convective heat transfer coefficient (W m2 K1) H: height of space (m) Icl: resistance to heat transfer k: turbulent kinetic energy (m/s) K: thermal conductivity of fluid (W m1 K1) L: Thermal load M: Metabolic rate P: pressure (Pa) Q: heat flux (W/m2) Ta: air temperature Tcl: clothing temperature Tr: radiant temperature Tu: turbulence intensity TN: ambient air temperature (K) ux: velocity component in x direction (m s1) v: velocity component in y direction (m s1) w: velocity component in z direction (m s1) yþ: normalized distance () Greek symbols b: thermal expansion coefficient (K1) r: fluid density (kg/m3) ε: rate of dissipation of kinetic energy u: specific rate of dissipation of kinetic energy n: kinematic fluid viscosity (m2/s) Acronyms CFD: computational fluid dynamics RANS: Reynolds averaged NaviereStokes SST: shear stress transport DTRM: discrete transport radiation model RTE: radiation transport equations PMV: predicted mean vote PPD: predicted percentage dissatisfied PD: percent dissatisfied