1、2009 ASHRAE 943This paper is based on findings resulting from ASHRAE Research Project RP-1321.ABSTRACTPrevious research on volatile organic compound (VOC)sorption by building materials shows that there are very few studies focusing on full-scale measurements and model vali-dations. It is not clear w
2、hether existing sink models based on small-scale chamber tests can be up-scaled to full-scale envi-ronments. In this study, both small-scale and full-scale cham-ber experiments were conducted to validate a diffusion sink model using sorption parameters obtained by small-scale chamber experiments. A
3、mixture of six compounds including ethylbenzene, decane, 1,2-diclorobenzene, undecane, benzal-dehyde and dodecane on a carpet was tested. The results indi-cate that the full-scale chamber itself has negligible sink effect. The sorption strength of six compounds on the carpet in the full-scale chambe
4、r has been found significant except ethylben-zene. Good agreement is observed between full-scale chamber data and predicted results by the diffusion sink model, indi-cating that the sorption parameters obtained by small-scale chamber can be up-scaled to full-scale conditions.INTRODUCTIONVolatile Org
5、anic Compounds (VOCs) constitute a signif-icant class of indoor air contaminants. There has been growing evidence that building materials can affect the transport and exposure of indoor VOCs by sorption. The re-emission of sorbed VOCs can dramatically elevate VOC concentrations in the indoor environ
6、ment for months or years after a source event (Tichenor et al. 1988; Berglund et al. 1988; Sparks et al. 1994). Accurate characterization of sorption of building mate-rials and the sorption impact on indoor air quality is important (Yang and Chen, 2001).Small-scale chamber tests are usually employed
7、 to char-acterize sorption capacity of a building material (An et al., 1999; Colombo et al., 1993; Huang et al., 2006.; Tichenor et al., 1991; Van der Wal et al., 1998; Won et al., 2000; Won et al., 2001; Yang et al., 2001; Zhang et al. 2001; Zhang et al., 2002). The utilization of a small-scale cha
8、mber is attractive because it has the potential to reduce the overall cost of VOC sink testing. Different sink models, including statistical models (Colombo et al., 1993; Tichenor et al., 1991) and mass transfer-based models (Axley et al., 1991; Little et al., 1996; Yang et al., 2001a) have been dev
9、eloped. However, the model parameters are usually obtained by fitting the model predic-tions with the small-scale chamber experimental data. It is not clear whether such models can be up-scaled to full-scale envi-ronments due to lack of rigorous validations under full-scale conditions.Compared to sm
10、all-scale chamber tests, conducting VOC sink tests using full-scale environmental chamber are more costly and time consuming, yet a necessary step toward model-ing indoor VOC transport in actual indoor environment. Pres-ently, there are very few studies focusing on large-scale measurements to provid
11、e information for scale-up of VOC sink models. Zhang et al. (1999) developed a method to measure sink effect of a 1942.3ft3(16.4 ft 13.1 ft 9.0 ft high) (55 m3, 5 m 4 m 2.75 m high) full-scale environmental chamber itself. Won et al. (2001a) conducted two types of scale-up experiments including the
12、exposure of three materials (carpet system, pad and virgin gypsum board) on three chemicals (cyclohexane, toluene and ethylbenzene) to validate a surface The Validation of a VOC Diffusion Sink ModelBased on Full-Scale Chamber TestQinqin Deng Jianshun S. Zhang, PhD Xudong Yang, PhDMember ASHRAE Membe
13、r ASHRAEQinqin Deng is a PhD student and Xudong Yang is Chang-Jiang professor in the Department of Building Science, Tsinghua University, China. Jianshun S. Zhang is a professor in the Department of Mechanical, Aerospace and Manufacturing Engineering, Syracuse University, Syra-cuse, NY.LO-09-089 (RP
14、-1321) 2009, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2009, vol. 115, part 2. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted witho
15、ut ASHRAEs prior written permission.944 ASHRAE Transactionssink model. A 367.3 ft3(10.4 m3) experimental chamber was used in the experiments. Relatively good agreement was observed between measurements and predictions by the surface sink model from the large-scale laboratory experi-ments. However, Z
16、hang et al. (2001) evaluated a surface sink model (linear Langmuir model) and a diffusion sink model for VOC sorption by different building materials, and pointed out that the diffusion sink model seems more suitable for the carpet.The main focus of this study is to validate a VOC diffusion sink mod
17、el applied to a carpet by developing a rigorous vali-dation procedure. Both small-scale and full-scale chamber experiments were conducted, while the full-scale chamber data were only used to validate the model predictions based on small-scale chamber test data (i.e., no curve fitting when comparing
18、the predicted and measured sink behavior in the full-scale chamber environments).METHODOLOGYFigure 1 illustrates the methodology applied in this study. First, VOC sink effect of a carpet is measured by small-scale chamber tests. The measured data are then analyzed by a diffu-sion sink model to obtai
19、n the sorption parameters. VOC sink tests using the same material and VOC are then conducted in the full-scale chamber. The measured data are compared with the predicted results based on sorption parameters determined from small-scale experiments to further validate the sink model. Table 1 shows bot
20、h the small- and full-scale test cases.Small-Scale Chamber TestSmall-scale chamber tests were conducted under a set of specific environmental conditions (e.g., 73.40.9F (230.5C), 50% 3% relative humidity (RH), and a ventila-tion rate of one air change per hour (ACH). Schematic of small-scale chamber
21、 sink test is shown in Figure 2. A stainless steel chamber with a volume of 50 liters as specified by ASTM D5116 (1997) was employed in the study. Each experiment included two periods: dynamic adsorption period and dynamic desorption period. During the dynamic adsorption period, constant VOC source
22、generated from the VOC gener-ator was carried by the conditioned, clean air from the condi-tioner (airflow, temperature and humidity controller) to the chamber containing the test specimen. The VOC concentra-tion in the chamber was monitored by analyzing air samples taken from the chamber exhaust at
23、 30 minutes, 1 hour, 1.5 hours, 2.0 hours, 2.5 hours, 3.0 hours, 6.0 hours from time zero during initial sampling. The initial sampling was used to capture the fast increase of concentration level in this period. Following the initial sampling, the sampling schedule was changed to moderate sampling
24、as the schedule of one sample at 12-24 hours intervals. Once the concentration at the cham-ber exhaust reached an apparent equilibrium (judged by the Figure 1 Illustration of the study methodology.ASHRAE Transactions 945outlet concentration getting close to the inlet concentration), the dynamic deso
25、rption period started whereby the VOC supply was stopped while the chamber was continuously flushed by the clean air, the samples in this period were taken as the same schedule as the adsorption period. Inlet concen-tration was also measured but at longer intervals during adsorption period, approxim
26、ately at 24 hours intervals during the adsorption period to measure the inlet concentration level and to make sure that the small-scale chamber test system is working properly. For desorption period, there was no inlet sampling because there was no VOC injection. The normal-ized concentration is obt
27、ained for analyzing sorption profile by dividing the concentration values in the chamber by the time-averaged, inlet concentration, which should be less than or equal to 1. However, due to measurement uncertainty and difficulty in controlling a constant inlet VOC source from the VOC permeation tubes
28、 (see below), the normalized value could be occasionally greater than 1 at a few sampling points.The sink effect of six compounds including ethylben-zene, 1,2-dichlorobenzene, decane, benzaldehyde, undecane, and dodecane on a carpet was studied. These compounds represent different types of VOCs incl
29、uding alkanes, aromat-ics, aldehydes and chlorine-substituted compounds and there is no known reaction among them. In addition, these compounds also cover a wide range of vapor pressures as shown in Table 2. For the carpet, the thickness of the backing was 0.1 in (0.003 m) and the thickness of the f
30、iber was 0.2 in (0.005 m). In order to minimize the edge effect, the specimen was placed in a stainless steel specimen holder 17.8 in by 8.0 in (0.452 m by 0.202 m) and the edges of the specimen were sealed by VOC-free aluminum tape (0.10 cm in width), giving an exposed surface of 17.7 in by 7.9 in
31、(0.45 m by 0.20 m). Each of these VOCs was generated by releasing it from a permeation tube placed in the temperature-controlled perme-ation chamber of the VOC generator, with a concentration about 0.5 mg /m3 during the tests. Previous study showed that the mixture effect on sorption is negligible u
32、nder such low concentrations (Won et al. 2001). Therefore, a mixture containing all above six compounds was released and tested simultaneously to speed up the experiments. The VOC concentrations at the chamber exhaust were measured using adsorbent tube air sampling followed by thermal desorption-gas
33、 chromatography/flame ionization detector (GC/FID) analysis referring to ASTM D6196 (1997). Before sample tubes of each test were analyzed by GC/FID, calibration curve were checked and updated. The VOC measurement uncer-tainty was estimated to be +/-15% relative standard deviation (RSD) for the comp
34、ounds tested (Zhang et al., 2001). However, it was found that the permeation rate of decane tube was unstable during the test, which resulted in the unstable inlet concentrations. Table 3 shows the inlet concentration data during the adsorption period in the small-scale chamber sink test.Chamber air
35、 flow rate test using tracer gas (sulfur hexa-fluoride), empty chamber sink test and background concen-tration test for small-scale chamber were conducted prior to VOC sink test on building material to examine the possible sink effect of the chamber itself. Results indicate that: (1) the airflow rat
36、e of the small-scale chamber was well controlled at 1.0 ACH (see exponential factor shown in Figure 3); (2) the small-scale chamber had insignificant sink effect for all the six compounds judged by insignificant deviations between the measured empty chamber data and simulation results based on no-si
37、nk assumption (see also Zhang et al. 2001). Empty cham-ber test result for decane is given in Figure 4. Detailed results for other compounds were reported by Zhang et al. (2001); (3) Table 1. Sink Tests CasesChamber Material Compounds Environmental Conditions ObjectiveSmall-ScaleChamberNo material S
38、ulfur hexafluoride73.40.9F (230.5C), 50%3% RHAir change rate testNo material DecaneBackground sink effect of chamberCarpet A mixture of six compounds* The validation of a sink modelFull-ScaleChamberNo material Sulfur hexafluoride73.41.8F (231.0C),50%5% RHAir change rate testNo material A mixture of
39、six compounds*Background sink effect of chamberCarpet A mixture of six compounds* The validation of a sink modelFigure 2 Schematic of the small-scale chamber sink test.946 ASHRAE Transactionsthe chamber has low background concentration (concentration of any individual target VOC is less than 2 mg /m
40、3).Full-Scale Chamber TestThe full-scale coupled indoor/outdoor environmental simulator, shown in Figure 5, has three major components: a 16 ft by 12 ft by 10 ft high (4.87 m by 3.66 m by 3.05 m) test chamber, a 6.5 ft by 12 ft by 10 ft high (1.98 m by 3.66 m by 3.05 m) outdoor climate chamber, and
41、a replaceable “separa-tion/test wall” assembly frame that is used to couple the two chambers. The test chamber is designed to simulate indoor environment while the climate chamber to simulate outdoor weather conditions. Both chambers and their respective heat-ing, ventilating, air conditioning (HVAC
42、) systems use stain-less steel interior surfaces and polytetrafluoroethylene gaskets to minimize pollutant emissions and adsorptions in the facil-ity. The HVAC systems for the test and climate chambers are both controlled with a direct digital control system, providing Table 2. Physicochemical Prope
43、rties of Tested VOCs*Compound Molecular Formula Molecular WeightBoiling Point,F (C)Vapor Pressure p0,mmHgEthylbenzene C8H10106.2 277.5 (136.4) 8.5821,2-dichlorobenzene C6H4Cl2147 357.1 (180.6) 1.185Decane C10H22142.3 345.7 (174.3) 1.246Undecane C11H22156.3 385.0 (196.1) 0.352Dodecane C12H26170.3 421
44、.7 (216.5) 0.115Benzaldehyde C7H6O 106.1 354.0 (178.9) 1.03*Yaws, C.L. 1999. Chemical Properties Handbook. New York: McGraw-Hill.Table 3. The Inlet Concentration Data During the Adsorption Period in theSmall-Scale Chamber Sink Test (mg/m3), Reference Condition (73.4F 23C, 50%)Time, h EB DEC BZA DCB
45、UND DOD0.00 0.747 0.448 0.371 0.471 0.366 0.241 24.00 0.760 0.460 0.383 0.493 0.370 0.260 67.00 0.763 0.451 0.371 0.502 0.353 0.251 Ethylbenzene (EB), 1,2-dichlorobenzene (DCB), Decane (DEC), Benzaldehyde (BZA), Undecane (UND), and dodecane (DOD)Figure 3 Airflow rate test of small-scale chamber.Figu
46、re 4 Empty small-scale chamber sink test.ASHRAE Transactions 947controls of temperature, relative humidity and pressure in both chambers. The details of the design of the simulator and the performance of its control system can be found in Herrmann and Zhang (2003).The carpet with 16 ft by 12 ft (4.8
47、7 m by 3.66 m) whose properties were tested by small-scale chamber measurements was placed in the full-scale test chamber, and gaps between the carpet and chamber walls were sealed with VOC-free alumi-num tape. The full-scale chamber test included three test peri-ods as shown in Figure 6. First, the
48、 chamber was preconditioned at 73.41.8F (231.0C) and 50% 5% RH. A known amount mixture of six liquid VOCs volatilized by hot plate was injected into the chamber (injection period). The VOC injection amount is calculated by the following equation:(1)where= estimated VOC injection mass in the full-sca
49、le chamber, mg= partition coefficient of the material, calculated based on small-scale chamber tests= estimated equilibrium concentration in the full-scale chamber, mg/ m3= volume of the material tested, m3= volume of full-scale chamber, m3Following the injection period, the supply and return fans were shut down and the chamber was maintained under the static condition for a period of 24 hours to allow the adsorption Figure 5 Photo of the full-scale stainless steel facility at Syracuse University. The left part is the climate chamber with the door shown. Th
