SYTTY
ADSORPTION, DESORPTION AND CHEMICAL REACTIONS IN THE PARTICULATE MATTER COLLECTED ON AIR FILTERS AND DUCTS
Project leader: Pertti Pasanen, University of Kuopio, Department
of Environmental Sciences P.O.Box 1627, FIN-70211 Kuopio, Finland, e-mail:
pertti.pasanen@uku.fi
(formerly) Pentti Kalliokoski, University of Kuopio, Department of
Environmental Sciences P.O.Box 1627, FIN-70211 Kuopio, Finland
| PUBLICATIONS |
| TIIVISTELMÄ SUOMEKSI |
Researchers:
Marko Hyttinen, tel. +358-17-163220, e-mail: Marko.Hyttinen@uku.fi
Pentti Kalliokoski, tel. +358-17-163150, e-mail: Pentti.Kalliokoski@uku.fi
Financing SYTTY organization: Tekes and The Ministry of Environment
Funding from SYTTY / Total funding of project (€): 199968
/ 199968
Person-months of work funded by SYTTY / Total person-months of work:
46 / 53
KEY WORDS: adsorption, desorption, dust, ozone, relative humidity
EXTENDED ABSTRACT
1 Introduction
Ventilation system is essential in maintaining good indoor air quality in modern buildings. However, as technical devices they need maintenance to function properly. Dirtiness may turn them into sources of odorous pollutants, which is the opposite to their planned influence (Fanger et al., 1988; Pejtersen et al., 1989). Air filter is the key component of the air handling system to prevent contamination of the system and indoor air. During the use, air filters remove particles from supply air and they become reservoirs of atmospheric dust and adsorbed VOCs, that generally makes the filters the strongest odor source in the ventilation system (Pejtersen et al., 1989; Finke and Fitzner, 1993).
The objective of this study was to find out the influence of adsorption and desorption phenomena in the particulate air filters on the quality of supply air. Chemical analyses, emission measurements and adsorption/desorption studies were conducted with dust sample collected from filters. Chemical reactions between ozone and pollutants adsorbed on particles on supply air filters and dusty duct surfaces were studied.
2 Methods
Adsorption/desorption phenomena were examined in a small scale dynamic test apparatus. The air used in the system was cleaned by activated carbon and its RH was adjusted by mixing dry air and air moistened by bubbling it through deionized water. RH was controlled by parallel flow controllers, which did not change the total flow, but only the ratio between humid and dry air. The air was the led through a small glass chamber which was loaded with the material to be tested. The tested materials, dust samples, were collected from loaded supply air filters in the Metropolitan area of Helsinki. Air samples were taken from both sides of the chamber onto Tenax GR adsorbent. Information of the filters is presented in table 1. Ozone measurements were carried out in laboratory and field experiments in Kuopio and Helsinki. Measurements were made by the Dasibi (Environmental Corp.) 1008-RS ozone analyzer. Concurrently, carbonyl compounds were collected on DNPH cartridges (Waters) and analyzed by HPLC technique.
Table 1. Information on the filters.
| Filter sample | Location | Operating time | Amount of air (m3) |
| G3 (HKI) | City of Helsinki | 10.11.2000-11.4.2001, 12h/day | 0,8*107 |
| F8 (HKI) | City of Helsinki | 10.11.2000-11.4.2001, 12h/day | 0,8*107 |
| G4 (JN) | City of Vantaa | 5.6.-11.12.2000, 12h/day | estimation 1,8*107 |
| F7 (JN) | City of Vantaaa | 5.6.-11.12.2000, 12h/day | estimation 1,8*107 |
| G3 (18t) | City of Vantaa | 18.5.-22.11.2000, 18h/day | 1,7*107 |
| F7 (18t) | City of Vantaa | 18.5.-22.11.2000, 18h/day | 1,7*107 |
Analysis
Total carbon and nitrogen contents of the dust samples were determined
by a CNH elemental analyzer. Elemental carbon (EC) and organic carbon (OC)
contents were analyzed by a thermal-optical carbon analyzer. Tenax GR,
filter and dust samples were analyzed by automatic thermodesorption device
(Perkin Elmer ATD400) connected to a gas chromatograph (HP-GC 6890) and
a mass spectrometer (HP-MSD 5973). Desorption conditions were: desorption
temperature 70oC/150oC for the filter and dust samples
and 250oC for the Tenax samples (holding time 10 min),
helium flow 50 ml/min, the cold trap temperature -30oC and secondary
desorption from the cold trap 200oC (1.5 min). The transfer
line temperature was 200oC and it was directly connected to
HP-5 silica capillary column (60m*0.25mm, film thickness 1.0 µm).
The temperature program was: 40oC hold for 1 min, increased
5oC/min to 220oC and then increased 20oC/min
to the final temperature of 250oC maintained for 11.5 minutes.
Identification of compounds was accomplished by retention times, standard
compounds, and GC-MS data library.
3 Results and discussion
The average carbon and nitrogen contents of the filter dusts were 17% and 1.2%. Content of elemental carbon was typically 3%. The carbon content was highest in the dust, which was collected from the pre-filter. Main emission products from air filter dust at the temperature of 70oC were aromatic hydrocarbons and in small extent aldehydes. Main emitted compounds at temperature of 150oC were carboxylic acids, alcohols, aldehydes, terpenes, phthalates, and aromatic hydrocarbons. Earlier studies have shown that intensity of odors from the filters was found to be strongest during the winter (Pasanen et al, 1995). In winter, the ambient particles derive largely from combustion and power production. The VOC emissions of the dust at the temperature of 150oC are presented in table. The main emission compounds are divided in to seven groups, the concentrations of which were measured as the sum of the areas of the individual compounds as toluene equivalents. VOC emissions of dust collected from the class G4 pre-filter was remarkable higher than those from the other filters. Emissions of dust samples collected from the downtown area of Helsinki (G3 Hki, F8 Hki) were relative low. This was probable due to the shorter service life of the filters compared to the others.
Table 2. VOC emission of the dust at 150oC (µg/g)
| Compounds | G3 (HKI) | F8 (HKI) | G4 (JN) | F7 (JN) | G3 (18t) | F7 (18t) |
| aldehydes (C5-C10) | 1,9 | 6,9 | 54,2 | 16,7 | 19,3 | 27,9 |
| ketones | 2,7 | 5,4 | 118,1 | 6,9 | 6,9 | 13,2 |
| terpenes | 1,3 | 7,7 | 2,2 | 0,1 | 2,0 | 5,0 |
| carboxylic acids (C2-C11) | 2,5 | 11,4 | 389,5 | 5,3 | 26,2 | 51,3 |
| aromatic hydrocarbons | 1,4 | 3,7 | 163,0 | 3,0 | 2,2 | 3,0 |
| aliphatic alcohols
(e.g. butanol, 2-ethyl-1-hexanol…) |
0,5 | 3,0 | 66,4 | 3,2 | 1,8 | 3,4 |
| others (e. g. furfural, pyridine, 2-methyl-furane. phthalanhydride, butylphthalate) | 4,0 | 10,5 | 153,8 | 80,8 | 75,2 | 8,3 |
| Total | 14,2 | 48,6 | 947,3 | 116,1 | 133,6 | 112,1 |
Adsorption and desorption properties of the dust accumulated on the filters were examined by using a small scale test apparatus with model compounds. Results indicated that adsorption of single compounds to the filter dust remained low in the test conditions, which were quite similar as the conditions in a real mechanical ventilation system. The specific surface area of the dust accumulated on the air filter appeared to be low, less than 1 m2/g. As the surface area is one of the most dominating factors for the adsorption capacity of a material (Ong and Lion, 1991; Lin et al., 1994; Ruiz et al., 1998), the capacity of dirty air filters to bind volatile air pollutants remains low. On the other hand, particles adsorb VOCs from the atmosphere already before entering the ventilation system (Weschler, 1984).
Desorption studies indicated that a constant relative humidity of air did not affect the rate of desorption. However, sudden increase in humidity increased substantially desorption of the model compounds. Similar results were obtained when the VOC emissions from loaded (used and dirty) EU7 filters were investigated without adding any model compounds. Downstream concentration of aromatic compounds remained higher than the upstream concentration throughout the whole test period. Emission was highest in the beginning of the experiment but decreased quite rapidly until the RH was changed: The raise of RH to 50% created a temporary increase in emission, which then again diminished until RH was increased. The same phenomenon appeared again when RH was raised to 80%. Similar temporary increase was detected in the emission of aldehydes with the raise of RH to 50%. However, no differences were observed before and after the filters when RH was raised further to 80%. The concentration level of aldehydes was also clearly lower, typically 2-7 µg m-3, than that of aromatic compounds (30-50µg m-3. There were no observations of compounds which would be of biogenic origin, such as terpenes. Also, the concentrations of aliphatic hydrocarbons remained low and rather constant on both sides of the filters throughout the test. The effect of relative humidity on the emissions of the aromatic compounds from filters is presented in figure1.
Figure 1. The effect of relative humidity on the emissions of aromatic compounds from EU7 filters. Air velocity was 0.17 m/s and temperature of air 20.7oC. X-axis is the time scale and the concentration ( µg m-3) of aromatic compounds is presented on the Y-axis as mean value of 20 minutes sampling period. Y-axis shows the concentration of aromatic compounds (µg m-3) before and after the test chamber at three different RH of air.
It seems that location and season are decisive to the quality of filter emission. In wintertime, filter dust contains mainly particles derived from traffic and combustion. An earlier test showed that, the intensity of odors from the filters was strongest during the winter (Pasanen et al., 1995). In the present experiments, emissions of aromatic compounds was detected. This result is quite expected because, air handling units, which facing the city streets with high traffic density collects effectively particles, which emits aromatic compounds. However, the influence of aromatic compounds to the intensity of odors from the filters remains unknown.
Ozone tests were carried out as small scale lab tests at University of Kuopio and at Helsinki University of Technology. In addition, ozone reduction caused by the ventilation system were studied at the field conditions in Kuopio and Helsinki by using similar measurement arrangements than in the laboratory. Studies revealed that ozone concentration did not change when air passed through an unused fiber glass filter. Thus, clean filter material do not react with ozone at ambient concentrations. On the other hand, the dust collected on the filters reacted with ozone. This effect was observed in 90% of the used filters, the ozone reduction ranged from lowest measurable difference of 0.5ppb up to 12ppb (4 to 26%). The reduction tended to increase with increasing ozone concentration. The highest reduction was obtained in a HVAC unit with three step filtration.
The measurements showed that the reduction increased with the increasing ozone concentration which is in agreement with reported influence of increased concentration on the reaction rate (Wecshler 2000). The changes in the concentrations of individual organic compounds were negligible except for formaldehyde and acetaldehyde which were produced in the reactions. Average ozone reduction in an oily duct of 41 meters (face velocity 3m/s) was only 1.1%. No reduction was observed in clean sheet duct. Measurements were made by using several ozone concentrations (20-150ppb). This confirmed earlier findings that air ventilation ducts do not work as sinks for ozone (Morrison et al. 1998).
4 Conclusions
Odor production is a common disadvantage of mechanical ventilation. The observed emission of odorous aldehydes and carboxylic acids during thermodesorption tests of the filter dust provides one explanation why air passed loaded ventilation filters has often been judged as stuffy in odor panel evaluations. The results were also consistent with previous odor observations in the sense that the emissions were higher from the pre-filter than from the fine filter of the same loaded ventilation system. However, the concentrations of odorous compounds observed were very low. An increase in the RH of air, increased the emission of many organic compounds: however, the effect seems to disappear quite rapidly. This at least partly explains why stuffy odors are often encountered when ventilation is turned on after nocturnal and weekend breaks. Thus, the avoidance of high RH of air on ventilation filters is important not only to prevent microbial growth but also to minimize the emission of VOCs from the dust collected on the filter. This investigation also revealed some interaction between dust particles and ozone in dusty air filters. Small amounts of formaldehyde were formed in these reactions. On the other hand, the reactions are insignificant on the duct surfaces.
5 References
Fanger, P.O., Lauridsen, J., Bluyssen, Clausen, G., 1988, Air pollution sources in offices and assembly halls, quantified by the olf unit. Energy and buildings 12, 1-6.
Finke, U., Fitzner, K., 1993, Ventilation and air-conditioning systems – investigations to the odour and possibilities of cleaning. In: Kalliokoski, P., Jantunen, M., Seppänen, O., (eds.), Proceedings of Indoor Air ’93. Indoor Air ’93, Helsinki, Vol.6, pp. 279-284.
Lin, T., Little, J.C., Nazaroff, W.W., 1994, Transport and sorption of volatile organic compunds and water vapor within dry soil grails, Environmental Science and Technology 28, 322-330
Morrison G.C. Nazaroff, W.W., Cano-Ruiz, J.A., Hodgson, A.T., Modera, M.P., (1998) Indoor air quality impacts of ventilation ducts: Ozone removal and emissions of volatile organic compounds. J. Air & Waste Manage. Assoc. 48:941-952.
Ong, S.K., Lion L.W., 1991, Effects of soil properties and moisture on the sorption of trichloroethylene vapor, Water Research 25, 29-36.
Pasanen P.O., Pasanen A-L., Kalliokoski P., Tuomainen A., 1995b, Prefilters and fine filters as a source of contaminants. In: Haghighat, F. (ed.), Proceedings of Indoor Air Quality, Ventilation and Energy Conservation in Buildings. Centre for Building Studies, Concordia University, Montreal, Canada, Vol 2., pp. 957-964.
Pejtersen, J., Bluyssen, P., Kondo, H., Clausen, G., Fanger, P.O., 1989, Air pollution sources in ventilation systems. In: Kulic, E., Todorovic, B., Novak, P. (eds.), Proceedings of CLIMA 2000, Air Conditioning Components and Systems, Yogoslavic Committee of Heating, Refrigerating and Air Conditioning and Union of Mechanical and Electrical Engineers and Technicians of Serbia, Beograd. Vol 3, pp.139-144.
Ruiz, J., Rafael, B., Murillo, M.B., 1998, Adsorption of different VOC onto soil minerals from gas phase: influence of mineral, type of VOC, and air humidity, Environmental Science and Technology 32, 1079-1084.
Weschler, C.J., 1984, Indoor-outdoor relationships for nonpolar organic
constituents of aerosol particles, Environmental Science and Technology,
18, 648-652.
Weschler C.J, 2000, Ozone in indoor environments: concentration and
chemistry, Indoor air, 10, 269-288.