1、4761 Measurements and Modeling of VOC Emissions from a Large Wall Assembly of Typical Wood-Framed Residential Houses Hui Li Student Member ASHRAE Miao Yang Student Member ASHRAE ABSTRACT A full-scale coupled indoor/outdoor environmental simulator (C-I/O-ES) was used to study the volatile organic com
2、pound (VOC) emissions from a typical residential wall assembly. The C-YO-ES has a stainless steel outdoor climate chamber, a stainless steel indoor environmental quality chamber, and a testheparation wall shared by both cham- bers. The wall assembly tested was a 3.66 m x 3.05 m high (123 x 103) sect
3、ion that was 0.2 m (8 in.) thick. It consisted of vinyl siding, vapor open weather barrier (house wrap), 0.012 m (% in.) oriented strandboard (OSB), 0.15 m (6 in.) mineral wool fiber bat9 insulation, polyethylene vapor barriel; 0.012 m (% in.) gypsum wallboard, and water-based paint. The wall assemb
4、ly was constructed in four stages: (1) OSB board with the wall frame; (2) adding the vinyl siding, house wrap, insulation, plastic vapor barriel; and gypsum wallboard; (3) adding two coats of paint; and (4) adding a window For each construction stage, VOC concentrations in both the climate and IEQ c
5、hambers were measured, and emission rates were calculated. The effects of the multi-layer system, air velocity, and air change rate on the contributions of each individual material in the assembly to the VOC concentrations in the IEQ chamber are discussed in this paper based on the measured data. In
6、 addition, a one-dimen- sional multi-layer VOC diffusion model was developed for the wall assembly. The parameters (diffusion coeficient, partition coeficient, and initial concentration) of VOC transport in individual materials were obtained from small- scale chamber tests and used in the model. Com
7、parison of simulation and experimental results show that the model prediction agrees well with the experimental data. J.S. Zhang Member ASHRAE Mikael Salonvaara Member ASHRAE I NTROD U CTION Indoor environmental quality (IEQ) can significantly affect human health, safety, comfort, and performance. I
8、EQ and the performance of the whole building envelope are inter- connected. A large number of VOCs are emitted from building materials. Most of the previous emission studies have focused on individual materials, especially interior surface materials, when estimating the effects of building structure
9、s on IEQ (Little et al. 1994; Guo et al. 1996; Zhang et al. 1999a, 1999b; Yang et al. 2001). However, surface materials are just a part of the potential contaminant sources in the indoor environment. Building envelopes as a whole and the materials in various layers of the building envelopes can be e
10、ven more significant factors affecting IEQ than surface materials when the long- term performance of the building is considered. This is because emissions from new interior surface building rnateri- als diminish over time relatively more quickly than emissions from the inner material layers. In orde
11、r to predict the effects of building envelopes on IEQ, it is necessary to understand the VOC transport in the building envelopes as well as the emis- sions of VOCs from individual material layers. The objectives of this study were to (1) determine the emissions of individual VOCs and total VOC (TVOC
12、) from different layers of a typical residential wall assembly and (2) develop a model to predict the VOC emissions and their impact on IEQ. EXPERIMENT Test Facilities Full-scale labora 3ry testing is a reliable method for studying contaminant sources and their transport in built envi- Hui Li and Mi
13、a0 Yang are research assistants, J.S. Zhang is an associate professor, and Mikael Salonvaara is a research scientist in the Department of Mechanical, Aerospace and Manufacturing Engineering, Syracuse University, Syracuse, N.Y. 210 02005 ASHRAE. l. Door 6. G,r D f hie- i1 Co3r 17. 5” Poly Insulation
14、2 Dota Coliection Ports 7. LTT li A.ter-otivr SUF: Aii- h!et IS. 11 Gouge Stainless Steel ?. Electricai Outlets 8. Dynariic Pressure Siidorcr 13. A ternati,ue Returr A.r Outlet 19. Light 3. Return Air Outlet 4. Window 8 E:ectz-icol d!ess 15. Ret%,rn Par Zutlet 21 Light 5. Supply Air Inlet 10 Window
15、13. 24 C-.ge Sxuinless Steel 9. Dota co;iPct:GT; POCIS 14. S,pr;iy Air- Idet 20. Diffuser Figure 1 The coupled indoor/outdoor environmental simulator (C-I/O-ES). ronmental systems. Several full-scale stainless chambers have been developed for material emission studies (Howard et al. 1995; Zhang et a
16、l. 1996; ASTM 2001). More recently, a coupled indoor/outdoor environmental simulator (C-I/O-ES) has been developed to study the combined heat, air, moisture and pollutant transport in building envelope systems as well as material emissions (Zhang et al. 2002; Hermann and Zhang 2003). The C-I/O-ES (F
17、igure 1) has three major components: a 4.87 m by 3.66 m by 3.05 m high (16 ft by 12 ft by 10 ft high) IEQ chamber, a 1.98 m by 3.66 m by 3.05 m high (6.5 ft by 12 ft by 10 ft) outdoor climate chamber, and a replaceable “separationhest wall” assembly frame that is used to couple the two chambers. Bot
18、h chambers and their respective HVAC systems use stainless steel interior surfaces and PTFE gaskets to minimize pollutant emissions and adsorptions in the facility. The HVAC systems for the IEQ and climate chambers are both controlled with a direct digital control (DDC) system, providing accurate co
19、ntrols of temperature, relative humidity, and pressure in both chambers. The operation mode for air exchange in the IEQ chamber has four options: (1) once-through (i.e., no recircu- lation), (2) recirculation, (3) normal (i.e., with partial recir- culation of air through the conditioning equipment),
20、 and (4) bypass (i.e., with recirculation air through a duct that bypasses the conditioning equipments). The operation mode for the climate chamber has two options: (1) recirculation and (2) normal. The “test wall” to study VOC emissions consists of a stainless steel frame with a wooden thermal brea
21、k. The section of the wall assembly tested had dimensions of 3.66 m (12 fi) Housewrap OS6 Fiberglass insulation lastic barrier vinyl sidiflrrlflr Figure 2 Structure of wall assembly of a typical residential house. wide, 0.2 m (8 in.) thick, and 3.05 m (10 ft) high; it consisted of 2.5 mm (1 O0 mil)
22、vinyl siding, 0.25 mm (10 mil) vapor open weather barrier (house wrap), 0.012 m (% in.) oriented strand- board (OSB), O. 15 m (6 in.) mineral wool (fiber batt) insula- tion, 0.25 m (10 mil) polyethylene vapor retarder, 0.012 m (% in.) gypsum wallboard, and water-based paint with a thick- ness of 0.0
23、2-0.05 mm (0.8-2 mil), as shown in Figure 2. The wall assembly was constructed by an experienced contractor. In order to determine emissions from different layers of the wall assembly, the assembly was constructed in four stages: (1) OSB board with the wall frame (wall configuration I); (2) adding t
24、he vinyl siding, house wrap, insulation, plastic vapor barrier, and gypsum wallboard (wall configuration II); (3) adding two coats of paints (wall configuration III); and (4) adding a window (wall configuration IV). Air samples were taken from both chambers by using adsorbent tubes, which were then
25、analyzed by a thermal desorption GUMS or GC/FID system. A ppbRAE was also used to monitor the total VOC concentration in the IEQ cham- ber during the test. ASHRAE Transactions: Research 21 1 O L1 Li Ln Y Q, chamber 4 , IEQ-makeup Q, C!a=O - IEQ-exhnurt Q, Ca-IEQ (i) Figure 3 Schematic of multi-layer
26、 diffusion model. Test Procedure All tests were conducted under the reference conditions of an air temperature of 23C and 50% relative humidity (RH) with the following procedure: 1. Establish stable test conditions (23“C, RH 50%, full recir- culation mode with 5 ACH total/supply/retum airflow, no ma
27、keup/exhaust for IEQ chamber and 26.7 ACH total airflow, no makeup for climate chamber) and check back- ground VOCs (take six air samples in each chamber within 24 hours). Wall configuration I (OSB with wall frame and wood studs) emission test: 2. (a) Construct wall configuration I. (b) Conduct a st
28、atic test: Establish and maintain refer- ence conditions in both chambers and take six air samples in each chamber within 20 hours. The sam- pling strategy was to take more samples during the early stage (when the VOC concentration changed more rapidly). The same strategy was used for all the follow
29、ing tests. (c) Conduct a dynamic test: Change the test conditions to normal mode (23“C, RH 50%, 0.5 ACH makeup, 5 ACH total airflow for IEQ chamber and 0.77 ACH makeup, 26.7 ACH total airflow for climate chamber) and take nine air samples within 48 hours. Wall configuration II (vinyl siding + house
30、wrap + OSB + insulation + plastic vapor barrier + gypsum wallboard) emission test: 3. (a) Construct wall configuration II. (b) Conduct a static test: Establish and maintain refer- ence conditions in both chambers and take six air samples within 18 hours. (c) Conduct a dynamic test: Change the test c
31、onditions to normal mode for both chambers and take 27 air samples within 144 hours. Wall configuration III (two coats of paints added to config- uration II) emission test: 4. (a) Construct wall configuration III. (b) Conduct phase I of a dynamic test: Establish and maintain the test conditions at n
32、ormal mode for cli- mate chamber and at once-through mode (5 ACH makeup, no return au) for IEQ chamber and take 56 air samples in two weeks. (c) Conduct phase II of the dynamic test: Reduce the IEQ makeup air to 2.5 ACH and take 28 air samples in one week. (d) Conduct phase III of the dynamic test:
33、Change test condition of IEQ chamber to normal mode Ca = curve-fitted VOC concentration in the chamber bulk E = voc emission rate, pg/(m2s; Q = makeup/exhaust airflow rate of the chamber, m3/s; t = elapsed time, s; V = volume of the chamber, m3. The measured VOC concentration data were curve-fitted
34、first, and then the curve-fitted data were used to calculate the emission rate. The experimental data (VOC concentrations in the chambers and emission rates of the wall assembly) were compared with the results from the numerical models. NUMERICAL MODEL Consider a multi-layer material source, both si
35、des of which are exposed to the air, as shown in Figure 3. Previous studies have indicated that the emission rates of VOCs from dry materials (e.g., OSB, gypsum board) are primarily controlled by the difision within the material (i.e., internal air, pg/m3; 21 2 ASH RAE Transactions: Research diffusi
36、on) (Roache et al. 1994). The emission process of wet materials (e.g., paints) appears to have three phases (Zhang et al. 1999b): (1) Phase I represents the period shortly after the material is applied but is still relatively wet. The VOC emis- sions are related to evaporation at the surface of the
37、material. (2) In phase II, the material dries as emission transition goes from an evaporation-dominant phase to an internal difision- controlled phase. (3) In phase III, the material becomes rela- tively dry and the dominant emission mechanism in this phase is the internal difision in the substrates
38、. The following one-dimensional multi-layer VOC trans- port model (Zhang et al. 1999b) was numerically solved in this study: Mass Transfer in an Individual Material dimensional diffusion model. The transport of VOCs in a material is described by a one- where C, = VOC concentration in solid materials
39、 at a certain Dm t = elapsed time, s Y = the location in the material, m location and time, pg/m3 = effective difision coefficient, m2/s MateriaIlAir Interface The so-called partition coefficient is used to present the storage capacity of VOCs in certain material for a given temperature under low VO
40、C concentration conditions: Cm = Kmaca (3) where Ca = VOC concentration in air at the material-air C, Kma = partition coefficient. MateriaIlMaterial Interface material 2, KI, can be expressed as interface, pg/m3; = VOC concentration in the material at the material-air interface, pg/m3; Partition coe
41、fficient between adjacent material 1 and (4) where Kmal, Kma2 are material-air partition coefficients for materials 1 and 2, respectively. The relationship between the concentrations of two adja- cent materials is presented by Equation 5. where, Cml, Cm2 are the VOC concentrations at the material- m
42、aterial interface in material 1 and material 2, respectively. On the other hand, the VOC transport between adjacent material layers can be expressed as inter-layer boundary conditions. At the material-material interface, where Dml, Dm2 are the effective difision coefficients in m2/s for the particul
43、ar VOC in the two layers, respectively. Mass Balance in the Chambers Assuming perfect mixing in the chamber and no VOCs in the makeup airflow (the background VOC concentration is negligible) to the chamber, the mass balance in the chamber can be described as Note that y = L, (Figure 3) is where the
44、material-air inter- face is located for the IEQ chamber side. For the climate cham- ber side, the material interface is located at y = O. The emission rate at the material-air interface is Initial and Boundary Conditions The boundary conditions for the multi-layer diffusion model are actually the ma
45、ss balance in the chambers as shown above in Equation 7. The initial condition for material layers of the wall assem- blies is c,gl, t = 0) = C,gl 9 (9) where Cmo is the initial concentration, pg/m3. assumed to be uniform in each material layer. At material AGE = O, the initial concentration is usua
46、lly For both the chambers, Ca = O) = Cao, (10) where Cao is the initial concentration in the chamber, pg/m3. Note that although emission from “wet” material (paint) was measured and discussed in this study, the modeling was limited to the dry materials only. Numerical Method Based on the above mathe
47、matical model, a code was developed in FORTRAN. A cell-center finite-volume implicit scheme was used in the code. The grid size was 0.1 mm to 0.2 mm (determined by D, and K, of the mate- rial), and the time step was ten seconds. ASHRAE Transactions: Research 21 3 Table I. UOC Transport Properties of
48、 Tested Building Materials (Yang et al. 2003) _ Material VOC K, Dm (m2/s) C, (pg/m3) OSB Octane 248 5.7510-“ 2.5010 Hexanal 3160 7.5510-l 1.8510 Tm-Hexanal I:l :1:,: 2.61;104 1 board Insulation Octane Hexanal 1168 1.2710- 1 .o0 C 2 0.80 5 8 e c 0.60 1 0.40 .- - m 0 0.20 z 0.00 . I (normalized by 4 2
49、 rng/rn3 ) x II (normalized by 12.1 mgh3 ) . 111 (normalized by 70.0 mg/rn3) p( O 10 20 30 40 Erne: t, h Table 2. Equations for Estimating D, K, (Yang et al. 2003) Figure 4 Total VOC concentration in IEQ chamber. Material Kmu D, (m2/s) OSB K, = 2440P-0.57 D, = 0.34Mw4.9152 Gypsum board K, = 45480P-.32 Dm = 74. 1MW-. Table 3. Comparison of Emission Rates (mg/m2h) of OSB between Small and Full-Scale Chamber Tests (Yang et al. 2003) Small Chamber Test (1.0 ACH; Loading Ratio of 1.634m-) Pentanal 0.017 0.025 Full Chamber Test (0.5 ACH; Loading Ratio of 0
