ASHRAE 4750-2005 Determination of Building Materials′ Transport Properties for Modeling VOC Emissions《测定建材？建模挥发性有机化合物排放的运动性》.pdf

《ASHRAE 4750-2005 Determination of Building Materials′ Transport Properties for Modeling VOC Emissions《测定建材？建模挥发性有机化合物排放的运动性》.pdf》由会员分享，可在线阅读，更多相关《ASHRAE 4750-2005 Determination of Building Materials′ Transport Properties for Modeling VOC Emissions《测定建材？建模挥发性有机化合物排放的运动性》.pdf（13页珍藏版）》请在麦多课文档分享上搜索。

1、4750 Determination of Building Materials Transport Pro pe rt es for Modeling VOC Emissions Miao Yang Student Member ASHRAE T.Q. Dang, PhD J.S. Zhang, PhD Member ASHRAE H. Li Student Member ASHRAE X.F. Gao ABSTRACT Material transport properties, such as difusion coefl- cients and partition coeficient

2、s, are necessary for modeling VOC emissions from building materials and their potential impact on indoor air quality. A modi3ed experimentalproce- dure for VOC emission testing was developed in this study to determine these properties. It consists of a static period (i.e., zero air exchange) followe

3、d by a dynamic period. The concen- trations measured at the end of the static period were used to represent the equilibrium concentration and, together with the data from the dynamic period, to determine the initial VOC content in the test specimen. An iterative numerical procedure with simulated an

4、nealing method was developedfor analyzing the experimental data to determine the partition and difusion coeflcients. The numerical procedure uses the measured data to estimate the initial value of the partition coejcient, and it was able togive a unique estimation of thepartiction anddifu- sion coef

5、icients from regresssion analysis of the measured data. Uncertainties of the experimental and numericalproce- dures were analyzed. Six building materials (vinyl siding, spunbonded olefn Tyvek TM, oriented strand board, jber batt insulation, gypsum wallboard, and interiorpaint) that are typically use

6、d in a wood-framed residential wall assembly were tested using a small stainless steel environmental cham- ber (50 L volume). VOCs emitted from each material were identified using thermal desorption GC/MS analyses of air samples collected in sorbent tubes. A thermal desorption-GC/ FID system was use

7、d to analyze the air samples to determine the VOC concentrations in the chamber as functions of time, and the VOC emission rates over time were also calculated. The numericalprocedure developed was used to determine the partition coefficient Zhang 1999). To use the model, one needs to know the parti

8、tion coefficient (K,) and diffusion coefficient (0,) of the materials for the VOC(s) of interest. A conventional chamber test method is characterized by a dynamic test only (Little et al. 1994). The proposed chamber test method is characterized by both static and dynamic tests. Equilibrium concentra

9、tion is obtained from the static test. The objectives of this study were to (I) develop an exper- imental procedure for determining K, and D, using small environmental chambers; (2) conduct uncertainty and sensi- tivity analyses to estimate the uncertainty of the method; and (3) determine the K, and

10、 Dm values for commonly used wall materials. Mia0 Yang and H. Li are graduate research assistants, J.S. Zhang is an associate professor, T.Q. Dang is a professor, and X.F. Gao is a former graduate research assistant in the Department of Mechanical, Aerospace and Manufacturing Engineering, Syracuse U

11、niversity, Syracuse, NY. aa 02005 ASHRAE. EXPERIMENTAL METHOD Test Facility A small environmental chamber test facility consisting of a clean air supply system, chamber assemblies, a humidity control system, a temperature control system, and a data acquisition system was used in this study (Figure 1

12、 from Zhang et al. 2002b). The small chambers, with dimensions of 0.5 m x 0.4 m x 0.25 m, were made of electro-polished 316 stainless steel. To create airtight sealing, Teflon gaskets (Gore-Tex joint sealant) were used to seal the covers of the chambers. Relative humid- ity for each chamber was meas

13、ured using an RH probe (HX94, accuracy +0.6“C), while temperatures at different locations inside the enclosure were monitored using thermocouples (accuracy +OSOC). The data acquisition system, a CR23x micrologger, was connected to thermocouples and humidity sensors to monitor the temperature and rel

14、ative humidity for each chamber. The SKC pocket-sampling pump and stainless steel sorbent tubes were used to collect samples in each test. The adsorbents used in this study were Carbopack C (60/80 mesh) and Carbopack B (60180 mesh) at a 2:l ratio (200 mg vs. 100 mg), while the GC was a Perkin-Elmer

15、Autosystem XL GC system with FID detection. The capillary column was a J RF, is the response factor of toluene (aredpg); Re, is the relative response factor, which is the ratio of the response factor of a selected chemical of interest against that of toluene. Therefore, M- - 1 e osampiing VOsampiing

16、Fict 2 2 2 = /(f7%) + (rtl%) + (f6%) + = I9.3%. Uncertainty in the impact of specimen variability Magee et al. (2003) conducted a series of small chamber tests of OSB to examine the degree of variability that can occur in the emissions of VOCs from a single building product. It was found that mill-m

17、ill differences were qual- itatively and quantitatively evident, as were differences between production dates, between individual panels with the same production date, and even between speci- mens taken from a single panel. Two samples of OSB from a single panel were tested and variation of equilib-

18、 rium concentration was about 10%. NUMERICAL MODEL AND DATA ANALYSES Governing Equations the following diffusion model (Zhang et al. 1999): The emissions from a dry material layer are described by For the material layer: For the material-air interface: For the air inside the chamber: Boundary condit

19、ions: S = -D rn .A!$ly=, at the material-air interfaceb = 1) (4) -D - - O at the bottom of the material layer (y = O) ay (5) where A = exposed surface area of the specimen, m2 Ca*(& CJt) = air phase concentration in the chamber air and at the air-material interface respectively. Assume that the conc

20、entration gradient in the boundary layer over the surface is negligible, pg /m3 ca(t)= c,*(t) = concentration in the supply air, pg /m3 = concentration in solid materials at a certain location 0.) and time (t), pg /m3 Cin Cm (y, t) ASHRAE Transactions: Research 91 Dm = diffusion coefficient, m21s Km

21、a = partition coefficient Q = supply airflow rate, m3/s S = emission rate as defined in Equation 4, pg 1s V = chamber volume, m3 Y = distance from bottom of the material layer, m Little et al. (1994) provided an analytical solution to this model that can predict emissions from individual materials,

22、but an analytical solution does not exist for multi-layer wall assemblies. In this study, we used a numerical approach to solve the equations so that the numerical simulation module can be readily used in predicting emissions from multi-layer wall assemblies with appropriate boundary conditions and

23、transport properties (i.e., K, and 0,) in the next phase of the study (see Li et al. 2003). Nondimensional Form of the Equations defined as The time scale for the diffusion inside the material may be 2 5 = 6 /D, where 6 is the thickness of the material, defined as the length scale. With this time an

24、d length scale, and choosing the initial concentration in the material as the reference concentration (Cmo), Equation 1 can be written in the following dimension- less form: where - Cm c, = - m i=-=- t t j=21 fi2/D, 6 Likewise, the VOC mass balance equation in the chamber can also be normalized. Thi

25、s equation constitutes the bound- ary condition for the above diffusion equation at = 1 (i.e., u=W where L=- -L 1 /6 (7) 1 Yes I I Output K,D, i Figure 3 Diagram of Iteration procedure. Numerical Solutions to the Governing Equations Applying a numerical technique to the mathematical model above, a c

26、omputer program was developed in FORTRAN with the cell-center implicit scheme. The grid size was 0.2 mm and the time step was 2 seconds. The program was also verified by comparing the calculation results against a validated existing program of a single-layer diffusion model (Gao 2001). Data Analysis

27、 Procedure In order to determine K, and D, from a measured concentration profile, Cexp(t), regression analysis using the diffusion model (Equation 1) is necessary. The iterationproce- dure for updating K, and D, is shown in Figure 3 and described as follows: 1. Obtain the measured and derived data i

28、ncluding the mate- rial thickness (6), material loading factor), air change rate (N), and the equilibrium concentration (C,). In the dynamic test, from 0-3 hours the concentration decreased quickly over time and the double exponential model -A, . f -B, , I C = Aoe +Boe was used in curve fitting. Aft

29、er three hours the concen- tration decreased slowly, and the power law model (C = a x fb)was applied in curve fitting. In order to have a smooth transition between these two models, the follow- 92 ASHRAE Transactions: Research 2. 3. Dm Initial Guess Final Km, Final D.(m2/s ing relation was also used

30、 in the curve fitting: at time t = 3 hour - dC - dC - at-1, = 3 hr atilt = 3 hr The initial guess of K, is C,/Ce, which was estimated as follows: (A)lE-10 (B)lE-11 (C) 1E-12 (C)lE-13 5.93E+03 5.8 1E+03 5.95E+03 5.95E+03 8.44E-13 8.73E- 13 8.43E-13 8.43E-13 Ca is the normalized concentration (Le., Ce

31、xp/Ces) and t is-the experimental time (i.e., 72 hours). Input initial guess of K, and D, into the difision model, and the difision model generates a concentration profile of Calculate the Ccal (t). then assign F, = Fold, Search the best K, D, to mini- mize the by using the simulated annealing metho

32、d, which is a stochastic computational technique derived from statisti- cal mechanics for finding globally minimum solutions to large optimization problems (Van Laarhoven 1987). Its major advantage over other methods is to avoid being trapped at local minima. It randomly chooses a trial point within

33、 the step length of initial value of D, and K, introducing a new set of D, (new) and K, (nav), Calculate the F, by using this Dm (,) If F nau Fold then assign O O f 1.00E-10 O O E -2 1.00E-11 .- p 1.00E-12 1.00E-13 OSB Bodalal, 1999 1 Gypsum Bodalal, 1999 1 i I l . D, = 0.802Mi5 76 Molecular Weiaht

34、Figure lob Difusion coeficient vs. molecular weight for OSB test (rough side). Plywood Bodalal. 1999 Figure I Od Partition coeficient vs. vapor pressure for OSB test (rough side). The correlations obtained in Figure 10 are useful to esti- mate D, and K, of certain VOC with known molecular weight and

35、 vapor pressure, which can be applied to simula- tions and predictions of VOC transport. SUMMARY AND CONCLUSIONS A modified experimental procedure for small environ- mental chamber testing of VOC emissions has been developed for determining the partition and diffusion coefficients of building materi

36、als, as well as the emission rates of the test specimen. It consists of static and dynamic test periods. An iterative numerical procedure based on the diffusion model and simulated annealing method was developed for analyzing the experimental data to determine the partition and diffusion coefficient

37、s. The numerical procedure uses the measured data to estimate the initial value of the partition coefficient and was 98 ASHRAE Transactions: Research Table 9. Kma and Dm of Hexanal and Pentanal from Different Materials Hexanal OSB Smooth Side OSB Rough Side Gypsum Board Plywood* Particle Board* D, (

38、m2/s) 7.55E- 12 2.88E- 1 3 8.94E-11 3.58E-11 7.42s-11 Kma 3160 15366 224 1845 2602 Pentanal OSB Smooth Side OSB Rough Side Gypsum Board Plywood* Particle Board* Dm(m2/s) 1.36E-11 1.35E- 12 9.00E-11 9.05E-11 3.66E-10 Kma 930 756 1385 680 1980 * Data from Bodalal(1YYY) Uncertainiy:+iS% able to give a

39、unique estimation of the partiction and difision coefficients from the regresssion analysis of the measured data. The VOC emissions of six building materials were tested and compared. OSB had the highest VOC emissions among the tested dry building materials. Difision and partition coef- ficients wer

40、e obtained and discussed for these materials. Correlations between D, and M, and between Km, and P were also developed and compared with previous findings. ACKNOWLEDGMENTS This study is supported by a research grant from U.S. EPA. I would also like to thank Dr. Lisa Cleckner of the NYIEQ Center for

41、her effort in editorial review and helpful comments. REFERENCES ASHRAE. 1997. 1997 ASHRAE Handbook-Fundamen- tals. Atlanta: American Society of Heating, Refiigerat- ing and Air-conditioning Engineers. Bodalal, A.S. 1999. A fundamental study of volatile organic emissions from building materials. Ph.D

42、 thesis, Carleton University, Ottawa, Ontario, Canada. Gao, X. 2001. Development and analysis of an experimental method and procedure for determining the diffusion and partition coefficients of VOCs emitted from drying building materials. Masters thesis, Syracuse University. Haghighat, E, C.S. Lee,

43、and W. Ghaky. 2002. Measurement of diffusion coefficients of VOCs for building materials: A review and development of a calculation procedure. Indoor Air 12:8 1-9 1. Hansson, P., and H. Stymne. 2000. VOC difision and absorption properties of indoor materials consequences for indoor air quality. Proc

44、eedings of Healthy Building 2000, Espoo, Finland, Vol. 4, pp. 15 1-1 56. Kirchner, S., J.R. Bradley, H.N. Knudsen, R. Meininghaus, D. Quenard, H. Sallee, and A. Saarinen. 1999. Sorption capacities and difision coefficients of indoor surface materials exposed to VOCs: Proposal of new test proce- dure

45、s. Indoor Air 1999, Edinburgh, Proceedings of the 8th International Conference on Indoor Air Quality and Climate, Vol. 1, pp. 430-435. Kline, S.J. 1985 The purposes of uncertainty analysis. Jour- nal ofFluid Engineering 107:153-160. Li, H., J.S. Zhang, and M. Yang. 2003. Measurements and modeling of

46、 VOC emissions from a large wall assembly of typical wood-framed residential houses. Proceedings of IAQ Problems and Engineering Solutions Confer- ence: Research Triangle Park, NC . Little, J.C, A.T. Hodgson, and A.J. Gadgil. 1994. Modeling emissions of volatile organic compounds from new car- pet.

47、Atmospheric Environment 28:227-234. Magee, R.J, D. Won, C.Y. Shaw, E. Lusztyk, and W. Yang. 2003. VOC emissions from building materials - the impact of specimen variability-A case study. Proceed- ing of IAQ Problems and Engineering Solutions Confer- ence: Research Triangle Park, NC. Meininghaus, R.,

48、 H.N. Knudsen, and L. Gunnarsen. 1998. Difision and sorption of volatile organic compounds in indoor surface materials. EPIC 98, Proceedings of the the Second European conference on Energy Perfor- mance and Indoor Climate in Buildings and the Third International Conference on Indoor Air Qualiy, knti

49、la- tion and Energy Conservation, Lyon, France, Vol. 1, pp. SalthammetT, C. Boehme, B. Meyer, and N. Siwinski.2003. Release of primary compounds and reaction products from oriented strand board (OSB). Proceedings of Healthy Building 2003 7th International Conference, Singapore. Schwope, A.D., W.J. Lyman, and RX. Reid. 1989. Method- ology for estimating the migration of additives and impurities from polymeric materials. Methods For Accessing Exposure to Chemical Substances, Vol 1 1. Tiffonnet, A.L., P. Blondeau, O. Amiri, and F. Allard. 2000. Asse