1、NA-04-8-1 Evaluation of Moisture Buffer Effects by Performing W hole-Bu Iding Si mu lat ions Carsten Rode, Ph.D. Nathan Mendes ABSTRACT The humidity of rooms and the moisture conditions of materials in the enclosure of buildings depend much on each other because of the moisture exchange that takes p
2、lace over the interior surfaces. These moisture influences also depend strongly on the thermal conditions of indoor spaces and enclo- sure elements of buildings. In turn, the moisture and humidity conditions have signiJicant impact on how buildings are oper- ated. In hot, humid climates, it may be d
3、esirable to keep the ventilation rates low in order to avoid too high indoor humid- ity, while in cold climates, ventilation can be used to keep the humidity low enough to ensure only a small risk of moisture damage in the building enclosure. In either case, the indoor humidity has a direct or indir
4、ect impact on the energyperfor- mance of the HVAC system of a building. To analyze this situation, it is todaypossible to benejtfrom some recent developments in integrated computational anal- ysis of the hygrothermalperformance ofwhole buildings. Such developments have led to models for whole buildi
5、ngs (the indoor climate, the enclosure, and the systems), which not only predict the thermal performance, such as in contemporary building energy simulation. A growing number of building energy simulation tools have added models for transient mois- ture migration. The paper gives examples of two suc
6、h recent developments and will highlight some calculation results that can be obtained. Finally, thepaper will mention some further developments and international collaboration for the near future, introducing a proposal recommendation for a common numerical test case. INTRODUCTION Indoor humidity i
7、s a result of vapor production caused by activities in indoor rooms and exchange of moisture with the Karl Grau surroundings, mainly through infiltration, ventilation, and other forms of air exchange. Indoor humidity also depends on the exchange of moisture with the building enclosure, as well as wi
8、th indoor furnishing. Most ofthis moisture exchange will be of a transient nature, such as the moisture exchange with indoor furnishing and materials on the interior surface of the building enclosure. Since the total surface area of materials in contact with the indoor environment can be significant
9、, and many materials may be hygroscopic, this exchange of mois- ture will act to moderate the indoor humidity variations seen in rooms under varying exposures. In addition, some moisture will also be exchanged with the exterior environment through various moisture transport processes that work acros
10、s the whole-building enclosure. Such transports can normally be seen as unimportant in comparison to the amounts exchanged by ventilation (Elmroth et al. 1997). The indoor humidity is one of the most important reasons for moisture accumulation in the building enclosure. Thus, there is a need to deve
11、lop good analytical techniques to eval- uate the integral moisture performance of the whole building, comprising the indoor environment and its enclosure. Recent attempts to develop such analytical techniques have been presented by Rode and Grau (2003), Mendes et al. (2003), Holm et al. (2001), and
12、Simonson et al. (2002). In addition, there is a need to make more experimental investigations in whole buildings, such as done by Simonson (2000). Realiza- tion that these needs exist has led to the formation of an inter- national research project in the framework of the International Energy Agency,
13、 which started at the end of 2003 and involves researchers from as many as i9 countries (Hens 2003). In cold climates ventilation of indoor rooms is maintained to, among other reasons, avoid excess indoor humidity levels, Carsten Rode is with the Technical University of Denmark. Nathan Mendes is wit
14、h the Pontifical Catholic University of Paran, Brazil. Karl Grau is with Danish Building and Urban Research. 02004 ASHRAE. 703 Since the indoor humidity loads vary-typically, with a daily cycle-an interesting question is whether proper utilization of the moisture buffering capacity of materials coul
15、d keep the indoor humidity at moderate levels and thereby limit the requirement for ventilation as a means to control the vapor content in periods when the rooms are occupied. Moisture buffering might also avoid very dry indoor humidity levels such as seen in cold climates in wintertime. In short, m
16、oisture buffering may eliminate the peaks and valleys in indoor humidity levels and thereby may contribute to an optimized indoor environment. There exists a lack ofknowledge about the extent to which such effects are worth pursuing (Rode et al. 2004). This paper will demonstrate the performance of
17、two different whole- building hygrothermal simulation programs. They are used to quantifi the daily moisture adsorptioddesorption for building enclosures with different materials, different HVAC system operation, and different outdoor climatic conditions. MOISTURE BUFFER EFFECT In a general way, moi
18、sture buffer capacity can be defined as a materials ability to reduce variations within anenclosure. For example, instead of relative humidity oscillating between say 40% and 80% due to indoor activities, proper use of mois- ture buffering material might limit variation between a range of 55% and 65
19、% RH or maybe even less variation. Moisture Buffer Definitions Moisture buffer capacity is an essential term for the topic of this paper. However, the term lacks a stringent definition. Some material properties that influence the buffer capacity are density, moisture capacity, and water vapor permea
20、bility. The density is of importance since a fixed volume of material with a high moisture capacity and a low density has only a small ability to contain water as opposed to the same volume of a material with a higher density. The moisture capacity is expressed by the gradient of the sorption curve,
21、 where the sorption curve gives the relationship between the equilibrium moisture content of a material and the relative humidity of its surround- ings. The sorption curve is not linear and, thus, the moisture capacity depends on the actual moisture level. Water vapor permeability is a material prop
22、erty that describes the rate of moisture transport by diffusion per unit area and vapor pressure difference through a unit thickness of material. Several different ways of defining moisture buffer capac- ity have been suggested. One way is analogous to thermal efu- sivity, which expresses a material
23、s capacity to absorb heat when exposed to a given thermal excitation (Hagentoft 2001). Inspired by this definition, a buffer effect can be described as a moisture effusivity, which derives from vapor permeability (“, m2/s), moisture capacity (5, kg/m3), and the saturation vapor concentration (us, kg
24、/m3). Moisture effusivity, b, : The unit for moisture emisivity is m/“. The moisture esivity is different from the well-known term moisture difusivity. 6, “s D, = - F 5W But the following relation can be established between the two: 6, b, - JDm Another measure of moisture buffer capacity is penetra-
25、 tion depth (E, m), which is also a combined parameter. It includes cycle time, water vapor permeability, saturation moisture content in the air (which is highly temperature dependent), and the moisture capacity. The penetration depth gives the active layer of a construction that is able to exchange
26、 moisture with its surroundings during a given cycle. The depth at which 37% of the variation at the surface can be registered is Moisture penetration, E : /F, where Dm 5 Padfield (1999) used the term available water during a given period to compare moisture buffer capacity of different materials. T
27、he available water is given by the product ofmois- ture capacity 5, and penetration depth E. The unit for available water is g/m2/day. A so-called Feuchtepufeerfunktion was introduced by researchers from the University of Essen, Germany, and an overview of its definition is given by Reick (2001). Th
28、e Feuchtepufferfunktion represents a mathematical paradigm to analyze the response of a material when subjected to a step change in surrounding moisture conditions. It is not a material property as such. Mitamura et al. (2001) introduced yet another way to express the buffer capacity. The weight of
29、a tested sample as a function of variation in ambient relative humidity was proposed. The drawback of this measure of buffer capacity is that it requires all materials to be tested experimentally for this specific parameter rather than adapt data that are available from other moisture property measu
30、rements. The various measures of moisture buffer capacity are hard to compare. They have different units and contain different parameters. It is not clear which parameters are the most = moisture disivity of the material, m2/s, and = period of the cycle, s. 784 ASHRAE Transactions: Symposia 1 O P O
31、X L Figure 1 Typical daily oscillation of moisture content for walls of high Bi, numbers. significant for buffer capacity. This topic was the issue of discussion of a recent international workshop (Rode et al. 2004), where illustrations were given of the different results that could be obtained usin
32、g some of the possible units for moisture buffer capacity. The workshop did not conclude which of the properties is best used as a figure of merit for moisture buffer capacity, but the moisture efhsivity may seem a good choice since it expresses the quantity of moisture absorbed or released from a s
33、urface when subjected to a mois- ture excitation. MateriaIlAir Interactions and the Moisture Buffer Effect Daily cycles of temperature and relative humidity induce energy and mass pulses into building envelope walls. These pulses affect the spatial distributions of temperature and mois- ture content
34、 in the walls. The affected region has a penetration depth, E, and oscillation amplitude, A(x). Thus, in an inner region, )m,ilvn,(t)-v,n()l (9) i=1 for the latent load. In Equations 8 and 9, the following defini- tions are used: A, = area of the ith surface, m2, h mass (hm, ds), Tn,i(t) = temperatu
35、re at the ith internal surface of the considered zone, K, L = vaporization latent heat, J/(kg.K), u = water vapor density, kg/m3. The temperature and vapor density are calculated by the combined heat and moisture transfer model based on the theory by Philip and deVries (1957), quoted by Mendes et al
36、. (2002a). This model is based on moisture content as a driving potential for moisture within materials. However, the moisture content may be quite different between different materials, even when they are in equilibrium with one another. Thus, the moisture content profile may experience a jump over
37、 the inter- face between different materials, and the mathematical treat- ment of this discontinuity in the moisture content profile is described by Mendes et al. (2002b). There is a strong coupling between the governing equa- tions. It implies the usefulness of applying an algorithm that is capable
38、 of solving ali equation sets simultaneously. Mendes and Philippi (2004) described the computational performance of the MTDMA (Multitridiagonal matrix algorithm), applied to the case of strongly coupled heat and moisture transfer in porous building materials. Engineers and researchers were motivated
39、 in the past to numerically decouple the governing equations because of the difference between the time scales for heat and moisture trans- = convection coefficients for heat (hc, W/m2.K) and fers. Nonetheless, besides the mathematical coupling between the heat and moisture terms in the conservation
40、 equations, the transport coefficients have a strong nonlinear dependence on both moisture content and temperature. In fact: For low moisture content, mass transfer is predomi- nantly in the vapor phase. Immediately after the liquid water has become a contin- uous phase, small changes in capillary p
41、otential may produce high variation in moisture content. Liquid transfer rates by capillarity will be greatly enhanced. This microscopic information about “liquid bridges” that speed up the moisture transport is embedded in the moisture diffisivity D,. When the moisture content is high, it is likely
42、 that high evaporation rates occur at the boundaries, promoting high gradients of moisture content and temperature and causing the tra,nsport coefficients to change consider- ably through the physical domain of the porous struc- ture. For the material chosen in this example (lime mortar), moisture t
43、ransport coefficients may rapidly change their magnitude by a factor of 100 or even more. Thermal conductivity may also have a substantial varia- tion on the order of 20%. These five facts might easily lead to problems with numerical divergence when the physical problem is mathe- matically decoupled
44、. For the water vapor balance, different contributions were considered: ventilation, infiltration, internal generation, porous walls, furniture, HVAC system, and respiration by occupants. In this way, the lumped formulation becomes where minf vent We, Winr 6 Jger Jsurf .IHVAc = vapor flow from HVAC
45、systems, kg/s. The water-vapor mass flow from the respiration of occu- pants is calculated, as shown in ASHRAE (1993). It takes into account the room air temperature, humidity ratio, and physical activity. The Brazilian building simulation model uses also the triangle model (Shewchuk 1996) that gene
46、rates a finite- element mesh on walls and the ground floor in order to = mass flow by infiltration, kg/s, = mass Bow by ventilation, kg/s, = external humidity ratio, kg waterkg dry air, = internal humidity ratio, kg waterkg dry air, = water vapor flow from the respiration of occupants, = internal wa
47、ter-vapor generation rate, kgs, = water vapor flow from porous surfaces (walls, kids, partitions, and hiture), kg/s, and 788 ASHRAE Transactions: Symposia precisely calculate shape factors and shadedsunny areas. The triangle model is fast, memory-efficient, and robust. It computes Delaunay triangula
48、tions and constrained Delaunay triangulations exactly. Guaranteed-quality meshes (having no small angles) are generated using Rupperts Delaunay refine- ment algorithm (see the triangulation in the bottom right side of Figure 9). SIMULATION RESULTS This section will illustrate some examples of use of
49、 the above-mentioned Danish and Brazilian whole-building hygrothermal simulation models. Test Case with the Danish Building Simulation Program In order to verify the moisture calculations with the Danish building simulation model and to illustrate the use of a practical test example for comparison and evaluation of whole-building hygrothermal models, the IEA BESTEST model (Judkoff and Neymark 1995) has been chosen to start. The IEA BESTEST represents an international effort to make intermodel comparisons of the thermal performance of some variations of a rather well
