1、NA-04-8-4 Predicting Indoor Temperature and Humidity Conditions Including Hygrothermal Interactions with the Building Envelope Andreas H. Holm, Ph.D. Hartwig M. Knzel, Ph.D. Member ASHRAE Klaus Sedlbauer, Ph.D. ABSTRACT The hygrothermal behavior of the building envelope affects the overallperformanc
2、e of a building. There are numer- ous tools for the simulation of the heat and moisture transfer in the building envelope and also whole building simulation tools for energy calculations. Howevei; working combinations of both models for practical application are just about to be developed. This pape
3、r describessuch a combinedmodel, which takes into account moisture sources and sinks inside a room, input from the envelope due to capillary action, and difusion and vapor absorption and desorption as a response to the exte- rior and interior climatic conditions as well as the well-known thermal par
4、ameters. By way of well-documented field exper- iments, the new model is validated and the moisture buffering capacity of the building envelope is determined. In the conclu- sions, the possible range of future applications of hygrother- mal building performance models is addressed and demands for fu
5、rther research are indicated. INTRODUCTION The heat and moisture behavior of the building envelope is an important aspect of the overall performance of a building. Today the hygrothermal transport phenomena through a build- ing enclosure exposed to natural climate conditions are well understood and
6、a number of models and computer codes have been developed and validated worldwide (Trechsel et al. 200 1). The same holds for thermal whole building simulations where a wide range of validated computer codes exists, e.g., ESP-r, TRNSYS, DOE-2, and EnergyPlus. However, very few models consider all hy
7、grothermal interactions between the indoor air and the building envelope in detail. A number of questions, which have gained importance lately, require a more accurate consideration of the hygrothemal processes in the building envelope, e.g.: How much ventilation and additional heating or cooling en
8、ergy is required to ensure hygienic indoor conditions when a building contains construction moisture or has been flooded? What happens to the building envelope when the indoor environment of a historic building is greatly changed, e.g., by turning it into a laundry or restaurant? How do different en
9、velope components react to fluctuat- ing indoor air conditions of buildings with temporary occupation? What humidity control strategies should be employed to preclude mold formation on the external and internal surfaces of the building envelope? Can vapor-absorbing finish materials help to save ener
10、gy and improve human comfort conditions? These questions can either be answered with the help of extensive experiments or by numerical simulations. In this paper a hygrothermal whole building simulation model and its experimental validation and an application under tropical climate conditions will b
11、e presented. The model takes into account the main hygrothermal effects, such as moisture sources and sinks inside aroom, moisture input from the enve- lope due to capillary action, diffusion and vapor absorption and desorption as a response to the exterior and interior climatic conditions, heat sou
12、rces and sinks inside the room, heat input from the envelope, the solar energy input through walls and windows, and hygrothermal sources and sinks due to natural or mechanical ventilation. Andreas H. Holm is head of the Department of Thermal Comfort and Climatic Effects, H.M. Kiinzel is head of the
13、Department of Hygro- thermics, and K. Sedlbauer is director, Fraunhofer IBP, Holzkirchen, Germany. 820 02004 ASHRAE. COMB IN IN G TH E RM AL BU I LD I N G SIM U LATI ON AND HYGROTHERMAL ENVELOPE CALCULATION As mentioned before there are a number of validated models for thermal building simulations a
14、s well as hygrother- mal envelope calculations used in building practice today. However, working combinations of these models are not yet available for the practitioner. In principle, this combination is achieved by coupling existing models of both types. Figure 1 shows the concept of such a combina
15、tion where balance equa- tions for the interior space and the different envelope parts have to be solved simultaneously. Recently the first real hygro- thermal simulation models have been developed (Karagiozis and Salonvaara 200 1; Rode et al. 2001), but so far only limited validation cases have bee
16、n reported. The model employed in this paper is called WUFI*+ (Holm et al. 2003) and is based on the hygrothermal envelope calculation model WUFI (Knzel 1994). The model for the hygrothermal envelope calculation, taking into account vapor difision, liquid flow, and thermal transport in porous materi
17、al, is based on the following equa- tions: Energy conservation: Muss conservation: where cp = relative humidity t = time, s 6 = temperature, K c = specific heat, J/kg.K w = moisture content, kgm3 psat = saturation vapor pressure, Pa h = thermal conductivity, W/(mK) H = total enthalpy, J/m3 D, = liqu
18、id disivity, m2/s p = vapor permeability, kg/(msPa) h, = latent heat of phase change, Jkg On the left-hand side of Equations 1 and 2 are the storage terms. The fluxes on the right-hand side in both equations depend on local temperature and humidity conditions. Equa- tions l and 2 must be solved for
19、every part of the envelope indi- vidually. Besides the exact definition of the assembly, including the material properties, the corresponding interior and exterior climatic boundary conditions are required. Usually the exterior boundary conditions are hardly affected by the building. However, the in
20、terior climatic conditions depend on several parameters, e.g., exterior climate, HVAC system, occupants behavior, humidity buffering of interior walls, and miture. The indoor air temperature i is linked to the heat fluxes into the room. This means that not only the heat flux through the envelope (tr
21、ansmission and solar input) is important but, in addition, internal thermal loads and the air exchange due to natural convection or HVAC systems must be taken into account. The energy balance can be described with the follow- ing equation. p.c. v.- d. = p.a.(e.-ei) dt JJ J (3) where p = density of t
22、he air, kgim3 a; = heat transfer coefficients, W/m2K 6, = exterior air temperature, K 6; = surface temperature, K 6, = indoor air temperature, K t = time, s Aj = surfacearea,mZ C = heat capacity of the air, J/kg K n = air change per hour, h- QSol solar input that leads directly to an increase of the
23、 air temperature or furniture, W Q, = internal gains, such as people, lights, and equipment, W Q, = heat fluxes gained or lost due to ventilation, W V = volume,m3 The humidity conditions in the room are a consequence of the moisture fluxes over the interior surfaces, the user- dependent moisture pro
24、duction rate, and the gains or loses due to air infiltration, natural or mechanical ventilation, as well as sources or sinks due to HVAC systems. = Figure 1 Coupling concept for the simultaneous treatment of the hygrothermal efects of interior heat and moisture loads, exterior climate, und transient
25、 behavior of envelope components. ASHRAE Transactions: Symposia 821 where Ca = absolute moisture ratio of the exterior air, kg/m3 Ci = absolute moisture ratio of the interior air, kg/m3 gwj = moisture flux from the interior surface into the W, = moisture production, kgh Wvent = moisture gains or los
26、es due to ventilation, kgh WHvAc = room, kg/(sm2) moisture gains or loses due to the HVAC system, kg/h. MODEL VALIDATION The thermal part of WUFI+ has already been validated by comparison with well-established building simulation tools (Holm et al. 2003); however, an important issue is the experimen
27、tal validation of these models under realistic but well-defined boundary conditions. Therefore, the following field test was designed for the validation of the humidity- related part of the model. Experimental Setup The experiments were carried out in a building erected on the IBP test site in the 1
28、980s, designed for energy investiga- tions, published in Knzel(l984). Two of the five rooms of this building are suitable for our purpose because they are identical. The floor plan of these rooms and the adjacent spaces is plotted in Figure 2. The rooms have a floor area of 20 m2 and a volume of 50
29、m3. They are well insulated (200 mm of polystyrene) toward the ground. In order to avoid.moisture flow to or from the ground, the floor has a vinyl covering. The exterior surfaces of the ceiling and partition walls face the conditioned space of the test building. The external walls consist of 240-mm
30、-hick brick masonry with 100 mm exterior insulation (ETICS also called EIFS). Walls and ceiling of the rooms are coated with 12 mm standard interior plaster. The double-glazed windows face south (U-factor: I. 1 W/m2K, total solar energy transmittance: 0.57, frame ratio: 30%). Special care has been t
31、aken to make the separate rooms as airtight as possible. Blower-door tests at 50 Pa resulted in air changes below 1 h-. For the new experiment the walls and ceiling of one room (test room) were rendered moisture inert by sealing them with aluminium foil while the interior surfaces of the other room
32、(reference room) were left as they are. Since the envelope of the test room has no sorption capac- ity due to the aluminium foil, it can be used to determine the moisture buffering effect of furniture and especially installed envelope components. In a second stage, wooden lining is applied over the
33、aluminium foiled interior surfaces in order to determine the moisture buffer capacity of wood. The room 600 Figure 2 Floor plan of the test room. followed by next test room Figure 3 Screen shot of the Internet-based visualization tool IMEDAS. with unpainted plastered walls represents traditional Eur
34、o- pean residences and serves as reference case. The rooms are equipped with calibrated heating, venti- lation, and moisture production systems as well as fans in order to avoid stratification. The indoor air temperature and humidity are measured at different levels above the floor. Temperature sens
35、ors and heat flux meters are also fixed to the interior surface of external walls. All values are measured on a five minute basis and can be analyzed with an Internet-based data acquisition and visualization tool called IMEDAS (see Figure 3). The first tests were conducted with a constant air change
36、 rate (ACH) of 0.5 h-, which is the hygienic minimum rate according to German regulations. The indoor air temperature in the middle of the room is kept above 20C. The moisture production is derived from an average moisture load (defined as the excess above outdoor air conditions) of 4 g/m3. This mea
37、ns the total amount ofwater dissipated in the room per day is 2.4 kg or 48 g/m3. In reality the production rate will not be constant over the whole day. Therefore, a basic production rate of 0.5 g/m3h is assumed with peaks in the morning and in the evening, Le., 8 g/m3h from 6 to 8 am. and 4 g/m3h f
38、rom 4 to 10 p.m. every day as shown in Figure 4. a22 ASHRAE Transactions: Symposia p 10 , 1 0.5 SI IIIIII 0.4 s fi g 0.2 1 0.3 Ij E! 0.1 0.0 +o 4 8 12 16 20 24 Hour of the day Figure 4 Diurnal moisture production pattern in the test rooms. Numerical Simulation The numerical simulations are carried o
39、ut with the version of WUFI+. The hygrothermal material parameters are taken from the WUFI 3.3 database. The moisture produc- tion and ventilation rate are the same as in the experiment. The outdoor climatic data (temperature, RH, solar radiation, etc.), which are continuously recorded at the meteor
40、ological station of the IBP, are introduced as hourly averages. Surface transfer coefficients are chosen according to Knzel (1994). As an initial condition, long-term dynamic equilibrium is assumed. Results The measured and calculated evolutions of the relative humidity in the empty test room during
41、 a day in January are plotted in Figure 5. Since the outdoor climate has been rather constant for some time and the moisture production and venti- lation pattern of the room are repeated every day, it is assumed that a dynamic equilibrium has evolved in the room. There- fore, the RH of the indoor ai
42、r at the end of the day is the same as at the beginning. There is perfect agreement between exper- iment and numerical simulation. The humidity fluctuations are greatest during the peak load in the morning where the indoor humidity rises from 35% to over 80%, which represents an increase of nearly 5
43、0% RH. Figure 6 shows the same results for the reference room with the plastered envelope. Again there is a rather good agreement between experiment and calculation with, however, a minor offset that is due to a slight difference in the conditions at the beginning of the day. It is assumed that the
44、moisture buffering effect of the envelope retards the dynamic equilibrium in the experiment. This leads to a slightly lower initial RH but does not influence the humidity fluctuations a great deal. The maximum increase in air humidity takes place again in the morning hours, but with Ca. 20% RH it is
45、 considerably lower than in the case of the aluminium foiled test room. This demonstrates the great influence of the vapor absorption capacity of the building envelope. loo 1 0- I I I -Ca Ic u la tion Measurement O 6 12 18 24 Hours of the day Figure 5 Simulated und measured evolution of the indoor a
46、ir humidity during a diurnal cycle in the test room coated with aluminum foil. 100 , I l l I n -Calculation =*o= iii e a s u re rn en t - O 6 12 18 24 Hours of the day Figure 6 Simulated and measured evolution of the indoor air humidity during a diurnal cycle in the reference room coated with a lime
47、-gypsum interior plaster ASHRAE Transactions: Symposia 823 -Calculation =-*- Measurement O0 06 12 18 24 Hours of the day Figure 7 Simulated and measured evolution of the indoor air humidity during a diurnal cycle in the test room when walls and ceiling were covered with wooden panels. After the init
48、ial tests the aluminium foiled test room was lined with wooden panels in order to increase its moisture buffering capacity. Apart from the outdoor climate, which was only slightly different, all other boundary conditions were left unchanged. The resulting indoor air humidity evolutions are plotted i
49、n Figure 7. The measured curve shows very limited humidity fluctuations. The maximum increase in the morning hours is less than 20%, which means that the wooden panels have a greater moisture buffering capacity than the unpainted interior plaster. However, this is not captured by the simulated curve, which clearly deviates from the measured one. Appar- ently the hygrothermal properties for wood in the database, which were determined by steady-state laboratory tests, do not represent the transient difision and absorption characteristics ofthe panel surfaces.
