|
Various authors have modeled the ventilated PV façade by evaluation of energy inputs and outputs through radiation, convection, conduction and power generation. Among those models, the simplest way is to use the single zone model and each element is represented by a node and connected together using the thermal network. The heat transfer through the elements is described using the U factor. Other studies have shown that modeling an AFW with a simple U-factor lacks the complexity to accurately model this type of façade (Saelens 2002).绿色建筑博客.dHH
{ p,_p!p
|'B:_ ] q0
Carla Balocco (2002) presented a simple model to study ventilated facades energy performance using a physical model to simulate the sample. A steady state energy balance was applied to a control volume which basic equations were solved by a finite element code with an iterative procedure. The simulated system consisted of an air channel, with insulated opaque wall on the side facing the air space and with an external spherical-granite covering panel of the façade. The heat transfer of the following aspects was examined:绿色建筑博客9w ?.E`v'xUU0O?7@:_
绿色建筑博客$L*C*c?
O8z ?X
(1) Convective and radiative exchange between the outside ambient and external covering, including solar radiation; (2) Radiative heat exchange between the two walls of the channel; 绿色建筑博客R)~1}+~2ta8}M
(3) Convective heat exchange between cavity walls and the circulating air mass 绿色建筑博客0j:ug:K7R8j
(4) Conduction through the walls 绿色建筑博客'H c3Dx;S:Y1^5^
(5) Heat exchange global coefficient between the opaque wall and the inside ambient that takes into account convective and radiative exchanges.绿色建筑博客,j H\ B%D
0t8r_ A.s7M _ V0
Li Mei and David Infield (2003) modeled the PV facades system within the TRNSYS simulation program. The building model comprised of three major components: the PV façade (PV panel, air gap and inner double glazing); the solar air collectors; and a TRNSYS single zone building model together with appropriate controller models. This paper outlined the main features of the complete model. To assess the thermal impact of the ventilated PV façade, heating and cooling loads for the building with and without façade were calculated. It was found that the cooling loads were marginally higher with the PV façade for all locations considered, whereas the impact of the façade on the heating load depends critically on location. From both measurement and simulation, it could be seen that the PV façade outlet air temperature reached around 50 ºC in summer and 40 ºC in winter. Twelve percent of heating energy could be saved using the pre-heated ventilation of the air for the building location in Barcelona in winter season. { x3D1rQ2gg0
} okYx8]0
The Solar Energy Laboratory at Concordia University (Charron and Athienitis 2003a, b; Liao and Athienitis 2005a, b) has developed three models to study the basic thermofluid processes and to optimize the BIPV/T system:绿色建筑博客:_$K[Y7} v o
1. A one-dimensional model that assumes isothermal surfaces but determines the exponential rise of the air temperature with height (Charron and Athienitis 2003a). 绿色建筑博客.B \%` k}
2. A two-dimensional model that divides the cavity into control volumes, allowing for the computation of the temperature distribution of the various components(Charron and Athienitis 2003b).绿色建筑博客Jzl? M7K_
3. A CFD model (part of this thesis) that is intended to study the airflow behaviour within the system, and to help determine appropriate convective heat transfer coefficients to use in models 1 and 2 (Liao and Athienitis 2005a, b).
p(X,~C#DMN!G;j0
绿色建筑博客Q K|(G]1V#`$Y };d O
The first two models were presented by Charron and Athienitis (2003) and the third CFD model is later described with details in this thesis. A simplified version of the one-dimensional model is presented in Figure 2.2 for Configuration 1 in Figure 1.1. This model was also used for quick validation studies. The two boundary surfaces (PV panel and back insulation) were assumed to be isothermal and were assigned an average temperature from experimental data. The air flow was analyzed using finite difference method to calculate the difference between the inlet and outlet air temperatures. Then the mean air temperature was used to calculate the heat transfer with boundary elements using the thermal network. 7E*O s ^ x y#`-x0
X V9y4ZVh g0
Figure 2.2: Thermal network model of ventilated façades with PV (assuming isothermal surfaces; node b indicates the back panel interior surface). 绿色建筑博客0TU,VXI
r:FB.P+c
绿色建筑博客+V`
m6y"L Na
An exponential variation is obtained for the air temperature as follows: 绿色建筑博客1g4w!hQt2U ~6[&E
/R'FrI J+Nd"@.\o0
where M = flow rate = V × A (V is average velocity and A is cross-sectional area) and W is width of façade. Note that the h is an average convective heat transfer coefficient for two cavity surfaces. The PV and back panel temperatures are obtained as follows: {#U@
B%[
EX ^6i4f0
#lW-R ^:t0
绿色建筑博客"sLA-|8@,o
where U represents conductance between the various nodes (Uo=A⋅ho, Ur=A⋅hr, Ua=Ub=A⋅h, and U3 is negligible). _(@m0^2v+gr0
h)a/H Y*G;o0{0
For most general cases, these thermal models can well predict the thermal performance and PV temperatures or insulation temperatures if proper convective heat transfer coefficients are used. The ways to determine the convective heat transfer can be found in many papers, however, they are for idealized flow and boundary conditions and they may not be suitable for our BIPV system. Thus intensive experimental validation work has to be done to find the suitable relationships. Second, the convective heat transfer coefficients used in the model are normally the average of the whole system. Since the system is working in the developing region of the cavity flow and relatively much stronger heat transfer is happening in the inlet region, the PV temperatures, insulation temperatures and the convective heat transfer coefficients all have a high gradient. Obviously, the average convective heat transfer can not present these local effects, thus detailed study of convective heat transfer in the air flow cavity is required. 1m5rN(R#O7N#MTi0 | 