1、4764 (RP-1071) Development of a Methodology to Quantify the Impact of Fenestration Systems on Human Thermal Comfort Jeet Sengupta Kirby S. Chapman, PhD Ali Keshavarz, PhD ABSTRACT This paper discusses the development of a methodology to quanti a review of the same appears in Brager and deDear (1998)
2、 and Watson and Chapman (2001) in general, but they do not focus specifically on quantifying the impact of windows. A joint study done by the Windows and Daylighting Group at Lawrence Berkeley National Laboratory (LBNL) and the Center for Environmental Design Research (CEDR) at the University of Cal
3、ifornia at Berkeley used a parametric approach to study windows and their effect on indoor comfort. Ten generic glazing systems, ranging from a single-pane window to high-performance windows, were examined for their thermal comfort impacts. The PMV was calculated for each combination of occupant/win
4、dow geometry, window surface temperatures, and clothing insulation. Results in terms of net PPD for each case were plotted to help visualize the trends for the dependence of thermal comfort on clothing and window type (Lyons et al. 1999). THERMAL COMFORT SIMULATION AND MODELING THE EFFECT OF WINDOWS
5、 It is understood from the above discussion that the impact of windows on human thermal comfort can be quantified if, for a room with windows, the air and mean radiant temperatures can be accurately calculated, since other quantifying comfort parameters depend on these two variables. The dry-bulb te
6、mperature is easily measured, while the mean radiant temperature can be calculated either by the classical method using surface temperatures and view factors as described in ASHRAE Fundamentals (200 I) or the radiant intensity method presented by Chapman and DeGreef (1997). The clas- sical method as
7、sumes the surfaces to be isothermal. Unfortu- nately, such is not the case for walls bearing windows. Because of different thermal and optical properties of the glazing, the glazing surface temperatures are distinctly different from the wall surface temperatures. In such cases the surface is subdi-
8、vided into smaller surfaces until the assumption of uniform temperatures is valid. This necessitates calculating a large number of view factors between the point under analysis and the surfaces as the number of surfaces increases. Further, since the glass is opaque in the infrared region, the radian
9、t energy in that region is reflected back into the room and this effect is not accounted for in the classical method. Also, in the case where solar radiation, which is short-wavelength radiation, shines through a window, the method fails as it does not consider window transmission and only considers
10、 the wall surface temperatures as boundary conditions. On the other hand, in the radiant intensity method, the mean radiant temperature is calculated by (Chapman and DeGreef 1997): (4) Since this equation provides a more generalized approach to calculating Tmrt than using the surrounding surface tem
11、per- atures, as given in the classical method and extensively vali- dated by DeGreef and Chapman (1 998), this method is used here to calculate the mean radiant temperature. However, this necessitates an accurate determination of the radiant intensity field. The easiest way to do so would be to solv
12、e the radiative transfer equation (RTE) (zisik 1977; Siegel and Howell 1981; Viskanta and Mengc 1987): + - (fi - Cl)Ih(R)dO Osh 4n I R For the case of a typical occupied room, the absorption and scattering coefficients can be assumed as zero and the equation reduces to 31, ar, 31, ax ay aZ p-+k-+q-
13、= o. The associated boundary condition in thex, y, andz direc- tions are given by (7) where p, = l-E*-T, (8) Each ofthe equations in equation set 7 has two terms. The first term represents the spectral emission from the surface, while the second term represents the reflected radiant inten- sity. Sev
14、eral models that engineers have developed to acquire the techniques of the RTE solution are documented in Siegel and Howell (1981) and Modest (1993). A review of the solu- tion techniques can be found in Chapman and Sengupta (2004) and DeGreef (1998). The technique adopted here is the discrete ordin
15、ates model that was first developed by Carlson and Lathrop (1968) in the neutron transport analysis. Jones and Chapman (1 994) used this model in ASHRAE Research Project 657 to develop an improved methodology, known as the BCAP methodology, to determine radiant exchange in an enclosure and incorpora
16、te the same in human comfort calcu- lations. The model, also used by Chapman and Zhang (1 995, ASHRAE Transactions: Research 241 1996) and Chapman et al. (1997), considers discrete directions and nodes and calculates the radiant intensity at each point and direction within the enclosure. The enclosu
17、re space is divided into a three-dimensional space of finite control volumes and Equation 5 is integrated over each three-dimensional control volume. The resulting equation in a discrete directionj is given by 4 e t t XYZ The control volume intensity on one side is assumed to be independent of the o
18、ther two directions. For example, the intensity on the x interface is not affected by the y and z direc- tions (Patankar 1980). The consequence of this procedure is that the intensity iw is uniform over the left surface of the control volume (Figure 3). Similarly, the intensity 4 is uniform over the
19、 bottom surface of the control volume. The equation then becomes PpAY (i; + Ax - iX) + CjAW + Ay - 4) (10) +qjAxAy(r,+,-i,) = O. This equation contains six interface intensities. Assuming the intensity profile across the control volume is linear, the intensity at the center of the control volume, po
20、intp, is (True- love 1988; Fiveland 1988) Ep = aix+Ax+(l-a)ix = a+Ay+(l -a) (1 1) - aiZ+,+(l-a)r, The interpolation factor a is set equal to 1 to avoid nega- tive intensities (Fiveland 1987, 1988), which are physically impossible and will yield unstable solutions. Substituting Equation 11 into Equat
21、ion 10 yields dAzAyi; + E;AzAx + $ Jamaluddin and Smith, 1988). The values for Ax, Ay, and Az are determined by the size of the control volume. The solution for Equation 12 is essen- tially an iterative solution solved in conjunction with appro- priate boundary conditions. The discretized form of th
22、e boundary conditions given by Equation 7 is: (13) Equation 12 is solved iteratively in association with boundary conditions given by Equation 13 to obtain the inten- sity in each of the j directions at pointp. These can be then used in Equation 4 to determine the mean radiant temperature. Note 242
23、ASHRAE Transactions: Research Table I. First Quadrant Values for Direction Cosines and Weighting Factors Ordinate Direction .uI 4j rll wi 1 2 3 4 that for a room with a glazing system, the boundary conditions need to be modified to incorporate the transmission through the window, The x direction bou
24、ndary condition can then be written as -0.2959 -0.9082 0.2959 0.5239 0.2959 -0.9082 0.2959 0.5236 -0.9082 -0.9082 0.2959 0.5236 -0.2959 -0.2959 0.9082 0.5236 where It was seen earlier that apart from the mean radiant temperature, the air temperature was the other quantity required to calculate the o
25、perative temperature and the PMV Though the air temperature can be easily measured using a dry-bulb thermometer, since a numerical technique was employed to calculate the mean radiant temperature, a numer- ical scheme can be easily adopted to calculate the air temper- ature as well. Such a numerical
26、 scheme would need a room air energy balance and a wall surface energy balance for each wall. For steady-state calculations the equations are: Ti)+ ai n.RI(R)dR- 19 cioT4 = O (17) The first term of Equation 16 represents convective losses to the bounding surfaces, and the second term represents the
27、air infiltration rate, where To is the temperature of the infiltrat- ing air. The first term of Equation 17 represents conduction through the wall where R, is the thermal resistance of the wall and the outer convective boundary layer. The second term is the convective heat flux between the inner sur
28、face and room air. The third term represents the incident radiant heat flux absorbed by the wall surface, while the fourth term is the radi- ant emission from the wall. Together these two terms equal the net radiant heat flux at the bounding surface. Equations 16 and 17 can be solved iteratively in
29、association with relevant bound- ary conditions. A finite difference method as described by Patankar (1980) was used for the numerical solution. Interest- ingly, the value ofl(Q) needed in Equation 17 can be obtained by solving the RTE using the discrete ordinates model delin- eated earlier. Further
30、more, solving Equations 16 and 17 will also give the surface temperature Ti for each wall surface element. BCAP Modified The Building Comfort Analysis Program (BCAP) meth- odology (Jones and Chapman 1994), briefly reviewed earlier, has been used with necessary modifications to suit the work presente
31、d here. Figure 4 gives the modified methodology. The boundary condition has been modified to include transmission through the glazing system, and they have been mathemati- cally described in Equation 14. The other enhancement has been to incorporate the surface properties directly from the Window 5.
32、1TM program and the ability to describe a frame around the glazing system. The last modification to the exist- ing BCAP has been the addition of PMV to quantiQ the ther- mal comfort along with the operative temperature. The PMV corresponding to the PPD of 10% is calculated at each node point in the
33、calculation domain, and the PMV distribution at a height of 1.25 m from the floor is plotted. REPRESENTATIVE CASE STUDY different cases: 1. 2. 3. 4. A test room 4 m x 4 m x 2.5 m high was analyzed for four The room without a fenestration system and heating system (Case D-R- 1 - 1) The room without a
34、 fenestration system but with a heating system (Case D-R-1-2) The room with a fenestration system but without a heating system (Case D-R-2-1) The room with a fenestration system and with a heating system (Case D-R-2-2) At first, calculations were done for a room without a fenestration system and wit
35、hout a heating system. The bound- ary conditions imposed were nighttime in the winter, requiring the outdoor temperature to be -18C. In the simulation, the exterior wall is the left wall of the room (refer to Figure 5a for wall designations). All other walls including the floor and the ceiling were
36、treated as interior walls and thus assumed bounded by other rooms. A temperature of 21C was applied to the interior surrounding spaces. The temperature limits correspond to the ASHRAE winter conditions. The U-factor of all bounding walls was set to 0.38 W/m2-K (0.067 Btu/h-f?- R), according to the p
37、roject specifications. Calculations were carried out for a relative air velocity of 0.1 m/s, which is ASHRAE Transactions: Research 243 I . Rcseribc the szca of the: ramn the filCs that wntcrrin DOE-2 output fbm Windaw SArn, and cvirometa input darn YES Figure 4 Modijied BCAP methodology (Chapman an
38、d Sengupta 2004). 244 ASHRAE Transactions: Research . 3 2 1 O 3 2 - RIGHT 1 O -. 3 ? 1 O Figure Sa The room and wall designation. Figure 5b The mesh. Caie D-R-2-1 case D-R.2-1 zx A 21 A 3 2 1 O Figure 6 Fenestration conjguration, slice for thermal Figure 7 Location of heaters. comfort signature. com
39、parable to natural convection, and relative humidity of 40%. A grid size of 0.25 m in all three coordinate directions was chosen as the acceptable maximum grid size (Figure 5b). The choice of number of grids in each coordinate direction was based on the grid independence check done by Zhang signatur
40、es and the thmd comfort Penetration. Figue 6 shows the fenestration configuration, and Figure 7 shows the location of heaters when radiant heat is added. Figure 8 gives the thermal comfort signatures wherein the comfort Parametem, the air temperature Tair, the mean radiant temperature Tmrt, the Oper
41、ative temperahire and the predicted mean vote (PMV) are shown as contour plots on a horizontal section across the room at a height of 1.25 m fi-om the floor, while Figure 12 (discussed below) gives the thermal Figure 8 shows that the dq-bulb temperature for the in-space air in the room with no fenes
42、tration is fairly uniform across the room at about 16C. The PMVplot in Figure 8 quantifies the (l994). The are shown in Ornfort comfort penetration. A close look at the Tair distribution in ASH RAE Transactions: Research 245 Figure 8 Case D-R-1-1, thermal comfort signatures (no fenestration and unco
43、nditioned). acceptability factor of the operative temperature in the room. Of note is that for the sedentary conditions with a clothing insulation value of 1.0 CIO and with a metabolic rate of 70 W/ m2, the operative temperature of around 16C does not make the room comfortable for at least 90% of th
44、e people occupying the room. In fact, as seen in Figure 8, no part of the room is in the acceptable criterion of PMV; being between -0.5 and +0.5 for a 10% PPD. This is interesting, since it shows that even for rooms without a fenestration system subjected to winter conditions, thermally comfortable
45、 areas do not exist for 90% of the people. Another interesting question to address before adding a fenestration system would be to evaluate the total power required for conditioning this space, i.e., a room without a 246 fenestration system with one wall subjected to outside temper- ature of -18C. T
46、he same space is conditioned by introducing radiant heat into the room. The heaters are placed inside the walls (see Figure 7 for locations), effectively making the walls heated walls. The right wall heater is of the same size as the right wall, while the ceiling heater, floor heater, front wall hea
47、ter, and back wall heater only extend for half of the walls. The choice of location of the heater probably has little signif- icance in a room without fenestration. However, the effort was to ensure that heating is introduced farthest from the wall exposed to the outside conditions. The reason was t
48、o ensure that the heat was introduced in a manner so as not to skew the radiant temperature within the space. This is especially critical once the fenestration is added to the walls exposed to outside ASHRAE Transactions: Research Figure 9 Case D-R-1-2, thermal comfort signatures (no fenestration an
49、d conditioned). temperatures. The prime objective in conditioning the space was to make it comfortable for at least 90% of the occupants at sedentary conditions involved primarily in office work. Since ASHRAE winter conditions were being simulated, the space was conditioned by heating. The space can be condi- tioned either by space heating, when simulating ASHRAE winter conditions, or by space cooling, when simulating ASHRAE summer conditions. Space conditioning is achieved by iterating to a solution such that the “thermostat setting” is 24C for summer or 2 1
