Professional Documents
Culture Documents
Bing “Beverly” Guo is a research engineer, Jianshun S. Zhang is a professor, Sunil Nair and Wenhao Chen are graduate research assistants,
and James Smith is a research design engineer, Building Energy and Environmental Systems Laboratory, Department of Mechanical and Aero-
space Engineering, Syracuse University, Syracuse, NY.
of media has also been widely used directly in practice in the measured to determine the following performance parameters
form of packed sorbent bed trays. of sorbent:
In order to improve and optimize the design of sorbent-
Ci – Co
based air-cleaning devices, performance tests for sorbent Removal efficiency, E (%) = ------------------- ⋅ 100
Ci
media need to be conducted for different categories of VOCs
typically found in indoor air. This paper briefly reviews Removal capacity, Cr (weight %) =
sorbent media selection criteria, describes a newly built test t ( Ci ) ( E ) ( MW ) ( Q ) ( 60 min/h )
system for sorbent media, and reports the test results for eleven ∫0 -----------------------------------------------------------------------
6
( 10 ) ( 24.414 ml/mol ) ( W )
dt
sorbent media and seven test VOCs. Analysis included test
repeatability, removal efficiency, removal capacity of where
sorbents, and effect of flow rate on the test results. Ci = inlet challenge gas concentration (ppm) as a
function of time
DESCRIPTION OF SORPTION MEDIA
SELECTED FOR EVALUATION Co = outlet challenge gas concentration (ppm) as a
function of time
The sorbent media selected are 11 products from four
E = removal efficiency as a function of time
manufacturers/suppliers. The selection reflected sorbent
types, intended use, and availability (in the USA and interna- MW = molecular weight of challenge gas
tionally). The sorbent media types for this project included Q = volumetric airflow rate (ml/min)
(1) activated carbon products—coal based, bituminous coal T = time (h)
based, coconut shell based, and bamboo wood based—and
W = media sample weight (g)
(2) activated aluminas with potassium permanganate products
(Table 1). Breakthrough time at 50% removal efficiency: Tb1/2
Half-life removal capacity: Cr1/2, i.e., Cr at t = 1/2 Te
EXPERIMENTAL METHOD
where
Principle Te = the time when the outlet concentration approaches
The test method used in this study is the same as that used equilibrium with the inlet concentration (Figure 2).
in VanOsdell and Sparks (1995) and that proposed in
ASHRAE Standard 145.1 (ASHRAE 2005), with some vari- Test System
ations in specific test system setup and test conditions. As An air-cleaning technology test system (ACTTS) was
shown in Figure 2, conditioned air passes through the sorbent developed in this study. The ACTTS resides in a temperature-
bed and VOC concentrations before and after the bed are controlled enclosure that is maintained at a negative pressure
Figure 5 Effect of flow rate on the removal efficiency and capacity: toluene tests.
Figure 6 Schematic of sorbent media holder. Figure 7 Preparation of sorbent media bed.
particles (e.g., < 2 mm) are tested, a thinner sorbent test bed effect of the initial fluctuations on the calculation of the equi-
might be used. librium gas-phase concentration, the time average of the
Therefore, in this study, we followed the ASHRAE measured concentrations during the last 12 hours before the
(2005) guide to conduct the tests under relatively high concen- start of the desorption period were used as the equilibrium (or
tration levels (10–100 ppm) for the purpose of comparing the “nominal”) inlet concentration in normalizing all the
performances of different sorbents under laboratory condi- measured concentrations, as presented Figure 8.
tions and for validating the proposed ASHRAE standard Figure 8 shows that the normalized outlet concentrations
(ASHRAE 2005). However, it should be noted that adsorption over time had a similar pattern among different tests: they
characteristics may differ between low (ppb) and high (ppm) increased over time and eventually reached equilibrium with
concentration levels. In general, at low concentrations there the inlet concentration. The rate of increase in the outlet
are sufficient adsorption sites for adsorbates (VOCs), and concentration and the time to reach the equilibrium condition
competitions among molecules for adsorption sites may be were, however, different for different sorbent media in the
neglected. The Henry Law for the adsorption isotherm would same test (i.e., for the same test VOC), which differentiates the
apply. At high concentrations, the increased number of mole- sorption characteristics of different sorbent media. For some
cules in the gas phase would compete for the adsorption sites. media and VOCs tested, a small breakthrough was also
A more complex isotherm model, such as the BET or Dubinin- observed at the very beginning of the test (i.e., at t = 0). The
Radushkevich (D-R) equations, would apply. The D-R equa- desorption period was started by stopping the injection of
tion has generally been considered the most acceptable for VOC to the inlet when all the sorbent channels reached the
activated carbons at high concentrations (Axley 1984). Data equilibrium condition. It is clear from the measured results
are currently lacking to determine whether it can be extended (Figure 8) that a certain amount of VOCs were desorbed.
to low concentration levels typically found indoors. In the next
phase of this research, we plan to conduct long-term testing Repeatability
under low VOC concentrations and develop/validate a method
for extrapolating the high concentration laboratory test results Some repeat tests were conducted for the same sorbent
to predict the performance under low concentrations typically media (#2) with toluene to examine how repeatable the results
encountered in indoor air. are among the different channels of the same test and between
tests (Figure 9). For the same test (channel #3 vs channel #1 in
RESULTS AND DISCUSSIONS the “5-10-05” test), where the inlet concentration and environ-
mental conditions were identical and media weight differed by
As examples, Figure 8 shows the measured data for five
only 2.3% (19.03 vs. 18.6 g), an excellent repeatability in the
sorbent media tested (M1, M2, M3, M4, and M11). It had been
concentration trajectory was observed (Figure 9). However,
difficult to set the inlet concentrations at the predetermined
some differences were observed between the different tests
exact level. The inlet concentrations had been allowed to differ
due to differences in inlet concentrations and media weight.
among the different VOCs as long as they were within the
The time needed for the outlet concentration to approach the
range of 10 to 100 ppm. This is considered short-term “ppm”
inlet concentration was longer for a lower inlet concentration
level testing for the purpose of comparing different sorbent
and a higher media weight, as expected.
media, as opposed to “ppb” level testing, which is more real-
istic for field application but can take much longer to
Removal Efficiency and Removal Capacity
complete. The inlet concentrations were stable overall in these
tests but had some small fluctuations during the entire test The performance of the 11 test media for removing tolu-
period, especially at the beginning of a test. To minimize the ene can be quantified from the calculated removal efficiencies
Ci Media†
VOC Name*
(ppm) M1 M2 M3 M4 M11
n-hexane 40 10.85 19.90 10.95 24.50 14.95
2-butanone 78 4.85 5.50 3.75 4.10 4.45
toluene 36 12.13 8.67 9.7 11.73 8.87
tetrachloroethylene 36 16.55 20.80 13.05 19.70 14.95
isobutanol 58 9.30 12.40 5.85 8.90 3.55
D-limonene 17 44.55 38.8 46.95 38.6 33.85
n-decane 34 21.05 24.4 22.95 18.2 26
*
VOC names are listed in the order of decreasing vapor pressure.
†
A bolded number represents the maximum Tb1/2 for the VOC in the corresponding row.
Table 4. Half-Life Time* Removal Efficiency (Er, % at t =1/2Te) for Selected 5 Media
mental runs, especially for different VOCs. To compare the A mechanistic model is needed to extrapolate test results
performance of different sorbent media, it is therefore more conducted at one concentration level to a different reference
concentration level for all different VOCs of interest so that the
convenient to conduct tests in a multi-channel system in which
performance of a sorbent media for different VOCs can be
different sorbent media can be evaluated under exactly the better distinguished. This is a topic that will be addressed in a
same inlet concentration levels. companion modeling paper (Nair et al. 2006).
ASHRAE. 2005. ASHRAE Standard 145.1P, Laboratory Test Young, R.T. 2003. Ad sorbent: Fundamentals and Applica-
Method of Assessing the Performance of Gas-Phase Air tions. New York: John Wiley & Sons.
Cleaning Media (draft as of February 2005). Atlanta: Zhang, Z., N. Gong, C.Y. Shaw, and L. Gao. 2000. Adsorp-
American Society of Heating, Refrigerating and Air- tion capacity of activated carbon for n-alkane VOCs.
Conditioning Engineers, Inc. IAQ International Conference, Raleigh, NC, July.
Axley, J.W. 1984. Tools for the analysis of gas-phase air- Zhao, X.S., Q. Ma, and G.Q.M. Lu. 1998. VOC removal:
cleaning systems in buildings. ASHRAE Transactions Comparison of MCM-41 with hydrophobic zeolites and
91(2):1130–45. activated carbon. Energy & Fuels 12:1051–54.