- Email: [email protected]
Formaldehyde emission—Comparison of different standard methods
ARTICLE IN PRESS
Atmospheric Environment 41 (2007) 3193–3202 www.elsevier.com/locate/atmosenv
Formaldehyde emission—Comparison of different standard methods Maria Risholm-Sundmana,, Annelise Larsenb, Ewa Vestina, Anders Weibulla a
Casco Adhesives AB, P.O. Box 11538, SE-10061 Stockholm, Sweden b IKEA of Sweden AB, 343 81 A¨lmhult, Sweden
Received 16 March 2006; received in revised form 9 October 2006; accepted 17 October 2006
Abstract The emission of formaldehyde is an important factor in the evaluation of the environmental and health effects of woodbased board materials. This article gives a comparison between commonly used European test methods: chamber method [EN 717-1, 2004. Wood-based panels—determination of formaldehyde release—Part 1: formaldehyde emission by the chamber method. European Standard, October 2004], gas analysis method [EN 717-2, 1994. Wood-based panels— determination of formaldehyde release—Part 2: formaldehyde release by the gas analysis method, European Standard, November 1994], flask method [EN 717-3, 1996. Wood-based panels—determination of formaldehyde release—Part 3: formaldehyde release by the flask method, European Standard, March 1996], perforator method [EN 120, 1993. Wood based panels—determination of formaldehyde content—extraction method called perforator method, European Standard, September 1993], Japanese test methods: desiccator methods [JIS A 1460, 2001. Building boards. Determination of formaldehyde emission—desiccator method, Japanese Industrial Standard, March 2001 and JAS MAFF 233, 2001] and small chamber method [JIS A 1901, 2003. Determination of the emission of volatile organic compounds and aldehydes for building products—small chamber method, Japanese Industrial Standard, January 2003], for solid wood, particleboard, plywood and medium density fiberboard. The variations between the results from different methods can partly be explained by differences in test conditions. Factors like edge sealing, conditioning of the sample before the test and test temperature have a large effect on the final emission result. The Japanese limit for F **** of 0.3 mg l1 (in desiccator) for particleboards was found to be equivalent to 0.04 mg m3 in the European chamber test and 2.8 mg per 100 g in the perforator test. The variations in inter-laboratory tests are much larger than in intra-laboratory tests; the coefficient of variation is 16% and 6.0% for the chamber method, 25% and 3.5% for the gas analysis method and 15% and 5.2% for the desiccator method. r 2006 Elsevier Ltd. All rights reserved. Keywords: Emission testing; Formaldehyde; Chamber; Building materials; Wood; Particleboard; Wood-based boards
1. Introduction
Corresponding author. Fax: +46 8 6428399.
E-mail address: [email protected] (M. Risholm-Sundman). 1352-2310/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2006.10.079
The emission of formaldehyde is an important factor in the evaluation of the environmental and health effects of wood-based board materials. Different test methods are used in different countries.
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
3194
In a global market it is of vital importance to be able to compare products with formaldehyde emission classes like E1 in Europe with F *** and F **** in Japan. It is also important that interlaboratory tests are performed to make it meaningful to compare results and possible to evaluate the variation that can be expected in the results obtained by different laboratories. In this article, comparisons of formaldehyde emissions measured with different test methods on solid wood, particleboard, plywood and medium density fiberboard (MDF) are presented. The impact of different test conditions, sample treatments, etc. are shown and discussed. 2. Standard test methods Determination of formaldehyde emission according to reference methods as the European chamber method EN717-1 requires, for example, chamber facilities and measurements until steady-state condition. There is, however, also a need for more
simple and less time-consuming standard test methods for production control and similar tasks. As some of the commonly used simple methods are performed at increased temperature (as EN717-2 and EN717-3) or includes extraction with solvents (as EN120), it cannot be expected that these methods always result in evaluations in compliance with results obtained by the emission chamber reference method. In Table 1, some main characteristics of the methods are given and they are also discussed in an article by Yu and Crump (1999). 2.1. Chamber methods The European chamber method, EN 717-1, is the reference method for the evaluation of the formaldehyde emission. The sample is placed in a chamber, normally 1 or 0.225 m3 in volume. The loading factor is 1 m2 m3 and the air exchange rate 1 h1. The temperature is held at 23 1C and the relative humidity (RH) at 45%. Formaldehyde released from the test pieces mixes with the air in
Table 1 Comparison of standard methods for the determination of formaldehyde emissions Method
Europe
Japan
Global a
EN 717-1 0.225, 1 or 412 m3 chamber EN 717-2 gas analysis 4 l chamber EN 717-3 500 mL flask EN 120 perforator JIS A 1901 20 l – 1 m3 chamber JIS A 1460 9–11 l desiccator JAS 233 9–11 l desiccator ISO/CD 12460 1 m3 chamber
Test sample
Test conditions
Size loading factor
Edge sealing (m open edge m2)
Temp/RH
Temp/RH
Air exchange/ hour
1 m2 m3
Partly (1.5 m m2)
23 1C/45%a
23 1C/45%
1
0.4 0.05 m
Yes
Not stated
60 1C/p 3%
15
0.025 0.025 m, 20 g 0.025 0.025 m, 110 g
No (80 m m2) No
Not stated
40 1C/~100%
No
Not stated
Toluene extraction at 110 1C
No
2.2 m2 m3
Yes
28 1C/50%
28 1C/50%
0.5
0.18 m2
No (27 m m2)
20 1C/65%
20 1C/0–80%b
No
0.18 m2
No (27 m m2)
Noc
1 m2 m3
Partly (1.5 m m2)
23 1C/50%a
Conditioning in the chamber, the values reported are steady-state values. See Fig. 4. c Stored at 20 1C for 1 day wrapped in plastic before testing. b
Conditioning
No 23 1C/50%
1
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
the chamber, and a specified volume of air is drawn from the chamber twice a day. The formaldehyde is absorbed in impinger flasks containing water and determined photometrically. The result of the test is given as the steady-state value in mg m3 or ppm (ml m3). In the Japanese 20 L chamber, JIS A 1901, the loading factor is normally 2.2 m2 m3 and the air exchange rate 0.5 h1. The temperature is 28 1C and the RH 50%. For this method, the result is given as the specific emission rate (EF) in mg m2 h1. When the loading factor and air exchange rate is 1, as in the European chamber, the concentration result in mg m3 is equal to the EF result in mg m2 h1. 2.2. Gas analysis The gas analysis method, EN 717-2, describes determination of accelerated formaldehyde release from wood-based panels. A test piece of known surface area is placed in a chamber with controlled temperature, RH, airflow and pressure. Formaldehyde released from the test piece mixes with the air in the chamber, and this air is drawn from the chamber. The formaldehyde is absorbed in water and determined photometrically. Particularly, the high temperature (60 1C), low RH (p3%) and high air exchange rate (15 h1) differ considerably from the chamber method and normal room conditions.
3195
based panels by extraction in a perforator. The formaldehyde is extracted from test pieces by means of boiling toluene and then transferred into water. The formaldehyde content of this aqueous solution is determined photometrically and expressed in weight per 100 g of dry board and is corrected for moisture content of 6.5%. In the standard it is noted that the correlation between perforator value and the formaldehyde emission depends on type of board material. Particularly, this difference between board materials in density, porosity and moisture can result in poor correlation between perforator values and emission results obtained in chamber methods. The perforator method is still a widespread method for production control. It is however often questioned from a work environment point of view due to use of toluene. 2.5. Desiccator method The two Japanese desiccator methods JIS A1460 and JAS MAFF 233 both describe determination of formaldehyde release from wood-based materials. Test pieces are placed in a desiccator containing a vessel with water. The formaldehyde released from the test pieces at 20 1C during 24 h is absorbed by the water and determined photometrically. As in the flask method, the RH is very high. 2.6. Comparison of different methods
2.3. Flask method The flask method, EN 717-3, describes determination of formaldehyde release from wood-based panels. Test pieces of known mass is suspended over water in a closed container at constant temperature. The formaldehyde released from the test pieces is absorbed by the water and determined photometrically. Particularly, the small test pieces with relatively large open edge area and large risk of sampling inhomogeneity, the container without ventilation, relatively high temperature (40 1C) and the high humidity generated in the container differ considerably from the chamber method and normal room conditions. 2.4. Perforator method The perforator method, EN 120, describes determination of formaldehyde content in wood-
There are large differences between European test methods as chamber method (EN 717-1), gas analysis method (EN 717-2), flask method (EN 717-3) and perforator method (EN 120) and Japanese test methods as desiccator (JIS A 1460; JAS 233) and the small chamber method (JIS A 1901). Test conditions like temperature, RH and air exchange rate vary substantially (see Table 1). Another important factor for differences in the emission results is the variation in required sample treatment, like (1) sealing of the edges, (2) sealing of the back and (3) conditioning before measurement. In the Japanese desiccator methods and the European flask method, the edges are not sealed even though they constitute a large part of the total area. Since the emissions in most cases are highest from the edges, this will have great impact on the results from these methods. Also considerable impact can be expected due to differences in conditioning; the gas analysis and JAS desiccator methods are run directly
ARTICLE IN PRESS 3196
M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
after removal of the plastic wrapping, while the other methods give values after conditioning of the samples in specified temperature and humidity. 3. Formaldehyde measurements included in this study 3.1. Inter-laboratory tests Samples of particleboard and MDF were sent to different laboratories for formaldehyde emission tests according to EN 717-1, EN 717-2 and JIS A 1460. The following comparisons were made: Chamber method, EN 717-1: Particleboards with thickness 10, 16 and 28 mm were used. Samples packed in plastic were sent to six laboratories in Europe and the tests were performed in both 225 l and 1 m3 chambers. The samples were cut at the laboratories and the edges partly sealed (1.5 m open edge m2). The results after 2 weeks in the chamber were compared (Winter Funch, 2004). Gas analysis method, EN 717-2: Particleboards with thickness 19 mm and MDF with thickness 10 and 22 mm, were used. Samples, cut to size and packed in plastic, were sent to 15 laboratories in Europe and Asia. The samples were analysed directly after opening the plastic and sealing the edges. Desiccator method, JIS A 1460: Two different particleboards, thickness 12 mm, and one MDF, thickness 12 mm were used. Samples were cut to size, packed in plastic and sent to eight laboratories in Europe and Asia. The samples were analysed after one week open conditioning at 20 1C, 65% RH. 3.2. Intra-laboratory tests For the determination of the reproducibility within one lab (tests made at different times), we made the following measurements for the different test methods: Chamber method, EN 717-1: Two MDF samples, which both gave constant concentrations in the chamber (mean values 0.19 and 0.15 mg m3, respectively), were measured 9 and 12 times respectively during a period of 14 days. A combined coefficient of variance (CV) was calculated. Gas analysis method, EN 717-2: Two calculations have been made. In the first, 12 test pieces were cut from a 12 mm MDF board and then wrapped in plastic. On six occasions during July 2004–January 2005, two of the pieces were run. A pooled standard deviation was calculated from these six duplicate results.
The second calculation was made from a total of 39 different samples (particleboards, MDF, plywood and edge glued panel samples) run during 2003. Also from these runs, a pooled standard deviation was calculated. Desiccator method, JIS A 1460: From a 12 mm MDF board 99 test pieces were cut and then stored open at 20 1C, 65% RH for 4 weeks. Nine pieces were put in each of eleven desiccators and the formaldehyde emission was measured according to the standard. After the measurement, the pieces were again stored open at 20 1C, 65% RH for 4 weeks and then measured again. This procedure was repeated four additional times and the CV for all measurements was calculated. 3.3. Comparison of different methods The formaldehyde emissions from solid wood, particleboards, plywood and MDF have been measured with different methods. The Japanese desiccator method JIS A 1460 was compared with European methods perforator EN 120 and chamber EN 717-1. Particleboards, thickness 12–19 mm, mostly of F **** quality were used, 21 samples for the EN 120 comparison and 8 for the EN 717-1 comparison. 3.4. Desiccator tests to determine RH in desiccators during test The RH was measured during the 24 h desiccator test using a capacitive humidity sensor, Testo 650, placed at the level of the sample. The humidity was measured in the desiccators as follows: 1. Reference, desiccator without sample. 2. Desiccator with an MDF sample conditioned according to JIS A 1460 at 20 1C, 65% RH for a week. The moisture content (MC) of the sample was 5.9% before conditioning and 8.2% after. 3. Desiccator with an interior plywood sample conditioned according to JAS MAFF 233 in plastic at 20 1C for 24 h. The moisture content of the sample was 7.6%. 3.5. Chamber tests to determine impact of temperature Four samples from the same MDF board were tested according to JIS A 1901, two at 23 1C and two at 28 1C in order to compare results from the
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
European and Japanese chamber tests. The emission was measured after both 7 and 14 days. 4. Results and discussion 4.1. Reproducibility in inter- and intra-laboratory tests The results from the inter-laboratory tests of methods EN 717-1, EN 717-2 and JIS A 1460 are given in Figs. 1–3. The CV, both within one laboratory and in the inter-laboratory tests, is presented in Table 2. Before the calculation of the CV for the results, each sample in the inter-laboratory tests were first normalized against the average of the results from all laboratories for the respective samples. As expected, the CV for the inter-laboratory tests is much higher than the intralaboratory CV. For the 39 board samples (of different kinds), the CV was calculated to 10%. This is much higher than the CV for the 12 pieces from the same sample, which is 3.5% and represents the spread in the method (excluding sample differences) (Fig. 4). 4.2. Comparison of different test methods The formaldehyde emission measured with different methods may be compared with the E1 limit of 0.1 ppm ( ¼ 0.124 mg m3) for chamber tests according to EN 717-1 and with the F **** and
3197
F *** limits of 0.3 and 0.5 mg l1 respectively for desiccator tests according to JIS A 1460 and JAS 233. If the results for the different samples are normalized against the values for particleboard E0, the diagram in Fig. 5 may be drawn. As can be seen, the other particleboard, E1, has approximately the same value (about 2) for each method. This shows that for the same kind of material, the methods show similar results; sample E1 emits twice as much as sample E0. Looking at the MDF sample, EN 717-1, EN 120 and JIS A 1460 show similar behaviour, this sample emits about three times as much as sample E0 according to all three methods. The gas analysis method, EN 717-2, on the other hand, gives a much more pronounced difference, while the flask method, EN 717-3, gives a value comparable to sample E1 (twice E0). The results for the plywood samples are similar for EN 717-1 and EN 717-2, while EN 717-3 gives a much higher value. This is probably at least partly explained by the huge difference in the ratio (open edge area)/(surface area) for the flask method compared to the other two. Solid wood gives low formaldehyde emission, close to the determination limit, for all methods. The result for the MDF sample, when tested in the European chamber according to EN 717-1 at 23 1C, was 0.10 mg m3. According to the conversion equation (ASTM D6007-96) this corresponds to 170 mg m3 at 28 1C. Since the loading factor and
Interlaboratory test, chamber method EN 717-1 10 mm PB, mean value 0.20 mg/m3
1.4
16 mm PB, mean value 0.18 mg/m3 28 mm PB, mean value 0.15 mg/m3
Relative response
1.2 1.0 0.8 0.6 0.4 0.2 0.0 1 - 1 m3
2 - 1 m3 3 - 1 m3 4 - 225 l 5- 225 l Laboratories - chamber size
6- 225 l
Fig. 1. Results from inter-laboratory tests performed according to EN 717-1 with MDF and particleboards. The results for each test have been normalized against the average of the results from all laboratories for the respective sample.
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
3198
Interlaboratory test, gas analysis method EN 717-2 19 mm PB, mean value 8.3 mg/m2,h
1.8
10 mm MDF, mean value 3.0 mg/m2,h 22 mm MDF, mean value 2.9 mg/m2,h
Relative response
1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 1
2
3
4
5
6
7 8 9 Laboratories
10
11
12
13
14
15
Fig. 2. Results from inter-laboratory tests performed according to EN 717-2 with MDF and particleboards. The results for each test have been normalized against the average of the results from all laboratories for the respective sample.
Interlaboratory test, desiccator method JIS A 1460 12 mm PB, mean value 0.24 mg/L 12 mm PB, mean value 0.64 mg/L 12 mm MDF, mean value 0.59 mg/L
1.40
Relative response
1.20 1.00 0.80 0.60 0.40 0.20 0.00 1
2
3
4 5 Laboratories
6
7
8
Fig. 3. Results from inter-laboratory tests performed according to JIS A 1460 with MDF and particleboards. The results for each test have been normalized against the average of the results from all laboratories for the respective sample.
the air exchange rate both are 1, this gives an emission rate of 170 mg m2 h1. The explanation for the substantially higher value obtained in the European chamber compared to the result in the Japanese chamber, 74 mg m2 h1, must be that the sample is run with partly open edges (1.5 m m2 open edge). It has been shown earlier (on parquet samples), that the emission from the edges is much higher than from the surface (Risholm-Sundman and Wallin, 1999).
The effect of conditioning may be seen in Table 3 from the results on solid pine obtained with the desiccator tests. In JAS 233, when the sample is wrapped in plastic until the test starts, the emission was measured to 0.18 mg l1. The conditioning for 1 week in JIS A 1460 at 20 1C 65% RH lowered the formaldehyde emission dramatically, to o0.1 (0.05) mg l1. The difference between our result in the gas analysis of solid pine, 0.3 mg m2 h1 and the value obtained by Meyer and Boehme (1997),
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
3199
Table 2 Reproducibility in inter- and intra-laboratory tests Method
Intra-lab No. of tests
EN 717-1 .225, 1 or 12 m3 chamber 21 EN 717-2 gas analysis 4 l chamber 12a 39b JIS A 1460 9–11 l desiccator 66
Inter-lab CV (%)
No. of labs
No. of samples CV (%) per lab
Outliers excluded
CV (%)
6.0 3.5a 10b 5.2
6 15
3 3
16 25
c
11 17
8
3
15
d
a
Test pieces cut from the same sample. Duplicate analyses of test pieces from 39 different board samples. c Result of one sample from one of six laboratories excluded (see Fig. 1). d All results from two of 15 laboratories excluded (see Fig. 2). b
100 90 80
RH (%)
70 60 50 40 JIS A 1460, MDF MC=8.2% JAS 233, Plywood MC=7.6% Empty desiccator
30 20 10 0 0
5
10
15
20
Hours Fig. 4. The RH in desiccators with samples of different moisture content (MC), as a function of time.
8 7 Normalised value
6 5
Value:16
EN 717-1 EN 717-2 EN 717-3 EN 120 JIS A 1460
4 3 2 1 0 Solid oak
Solid pine
PB E0
PB E1
Plywood int
Plywood ext
MDF
Fig. 5. Comparison of the results from different test methods on different samples. The results have been normalized against the particleboard (PB) E0 value.
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
3200
Table 3 Formaldehyde emission measured according to different standard methods Material
Solid wood
Oak MC ¼ 8% Pine MC ¼ 8%
Thickness
Chamber
Gas analysis
Flask
Perforator
Chamber
Desiccator
(mm)
EN 717-1 (mg m3)
EN 717-2 (mg m2 h1)
EN 717-3 (mg kg1)
EN 120 (mg per 100 g)
JIS A 1901 (mg m2 h1)
JIS A 1460 (mg l1)
JAS 233 (mg l1)
20
0.005a
0.05a
0.06a
0.19a
o0.1 (0.00)c
o0.1 (0.03)c
20
0.006a
0.3b
0.16a
0.23a
o0.1 (0.05)c
0.18
a
Particle board
Plywood
MDF
E0 MC ¼ 6.8% E1 MC ¼ 6.0% Interior MC ¼ 8.6% Exterior MC ¼ 6.2% MC ¼ 5.9%
12
0.03
0.09 0.8
2
2–3
0.2
12
0.07
2
4
4.6
0.5
0.20
5.5
15
0.01
0.2
1.4
12
0.10
4.4
3.5
9.5
32
4.7 0.3 7.5
74 (43 at 23 1C)
0.6
a
Value from Meyer and Boehme (1997). No conditioning. c Analysed value, below the determination limit of 0.1 mg l1. b
0.9
mg L-1 y = 6.8561x + 0.0463
0.8
R2 = 0.7294
JIS A 1460
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
EN 717-1
0.09
0.1
mg m-3
Fig. 6. Correlation between JIS A 1460 and EN 717-1 for particleboards, thickness 12–19 mm.
0.09 mg m2 h1 may also be an effect of different conditioning. The correlation between the Japanese desiccator method JIS A 1460 and the European methods EN 717-1 (chamber) and EN 120 (perforator) was not convincing, the R2 values obtained were 0.66 and 0.73 (see Figs. 6 and 7). The values for the European methods that correspond to F ****, 0.3 mg l1, are given in Table 4. These values agree quite well with values reported earlier (Bulian et al., 2004) for particleboards in the range of 0.01–0.1 mg m3.
4.3. Desiccator tests—RH The results from the humidity measurements in the desiccators show that wood-based samples will lower the RH from 100% to about 60–80% (see Fig. 4). 4.4. Chamber tests—temperature The MDF sample was tested according to JIS A 1901 at both 23 and 28 1C. The result obtained at 23 1C,
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
3201
mg L-1
JIS A 1460
0.9 0.8
y = 0.0825x + 0.0816
0.7
R2 = 0.7559
0.6 0.5 0.4 0.3 0.2 0.1 0 0
1
2
4
3
5
6
7
Perforator, EN 120
8 mg / 100 g
Fig. 7. Correlation between JIS A 1460 and EN 120 for particleboards, thickness 12–19 mm.
Table 4 Values in chamber test EN 717-1 and perforator test EN 120 corresponding to the F **** limit, 0.3 mg l1, in JIS A 1460
EN 717-1 (mg m3) EN 120 (mg per 100 g)
Result corresponding to F **** limit
Literature result (Bulian et al., 2004)
0.037 2.8
0.04 2.4
43.0 mg m2 h1, may be converted to 28 1C by using the conversion equation, C ¼ C0 e9799(1/t1/t0), according to ASTM D6007-96. This gives a value of 75, which agrees very well with the measured value 74, given in Table 3. 5. Conclusion Comparisons of different formaldehyde emission methods show that: The variations can be explained by differences in the test conditions. Factors like edge sealing, conditioning of the sample before the test and test temperature have a large effect on the final emission result. The Japanese limit for F **** of 0.3 mg l1 (in desiccator) for particleboards was found to be equivalent to 0.04 mg m3 in the European chamber test and 2.8 mg per 100 g in the perforator test. The correlation between the desiccator JIS A 1460 and the chamber and perforator methods respectively is, however, not convincing ðR2 ¼ 0:7Þ. The variations in inter-laboratory tests with chamber method EN 717-1 (CV ¼ 16%), gas
analysis EN 717-2 (CV ¼ 25%) and desiccator JIS A 1460 (CV ¼ 15%) are much higher than in intralaboratory tests. Different standards in different countries make it difficult for the manufacturer of wood-based board materials to be able to comply with regulations in different countries. There is a need for international harmonization. Standardization work within ISO should be encouraged. References ASTM D6007-02, 2002. Standard test method for determining formaldehyde concentration in air from wood products using small-scale chamber. American Standard, April 2002. Bulian et al., 2004. Formaldehyde testing of wood-based panels: correlations between European and Japanese test methods. In: Wood-based Panel Symposium in Hannover, September 2004. EN 120, 1993. Wood-based panels—determination of formaldehyde content—extraction method called perforator method. European Standard, September 1993. EN 717-1, 2004. Wood-based panels—determination of formaldehyde release—Part 1: formaldehyde emission by the chamber method. European Standard, October 2004.
ARTICLE IN PRESS 3202
M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
EN 717-2, 1994. Wood-based panels—determination of formaldehyde release—Part 2: formaldehyde release by the gas analysis method. European Standard, November 1994. EN 717-3, 1996. Wood-based panels—determination of formaldehyde release—Part 3: formaldehyde release by the flask method. European Standard, March 1996. ISO/DIS 12460, 2005. Wood-based panels—determination of formaldehyde release—formaldehyde emission by the 1 m3 chamber method. Draft International Standard, January 2005. JAS 233, 2003. Japanese Agricultural Standard for plywood, Japanese Agriculture Standard, February 2003. JIS A 1901, 2003. Determination of the emission of volatile organic compounds and aldehydes for building products—small chamber method. Japanese Industrial Standard, January 2003.
JIS A 1460, 2001. Building boards. Determination of formaldehyde emission—desiccator method. Japanese Industrial Standard, March 2001. Meyer, B., Boehme, C., 1997. Formaldehyde emission from solid wood. Forest Products Journal 47 (5), 45–48. Risholm-Sundman, M., Wallin, N., 1999. Comparison of different laboratory methods for determining the formaldehyde emission from three-layer parquet floors. Holz als Rohund Werkstoff 57, 319–324. Winter Funch, L., 2004. Danish Technological Institute. Report RRT on prEN 717-1, 2004.08.25. Yu, C.W.F., Crump, D.R., 1999. Testing for formaldehyde emission from wood-based products—a review. Indoor Built Environment 8, 280–286.
Atmospheric Environment 41 (2007) 3193–3202 www.elsevier.com/locate/atmosenv
Formaldehyde emission—Comparison of different standard methods Maria Risholm-Sundmana,, Annelise Larsenb, Ewa Vestina, Anders Weibulla a
Casco Adhesives AB, P.O. Box 11538, SE-10061 Stockholm, Sweden b IKEA of Sweden AB, 343 81 A¨lmhult, Sweden
Received 16 March 2006; received in revised form 9 October 2006; accepted 17 October 2006
Abstract The emission of formaldehyde is an important factor in the evaluation of the environmental and health effects of woodbased board materials. This article gives a comparison between commonly used European test methods: chamber method [EN 717-1, 2004. Wood-based panels—determination of formaldehyde release—Part 1: formaldehyde emission by the chamber method. European Standard, October 2004], gas analysis method [EN 717-2, 1994. Wood-based panels— determination of formaldehyde release—Part 2: formaldehyde release by the gas analysis method, European Standard, November 1994], flask method [EN 717-3, 1996. Wood-based panels—determination of formaldehyde release—Part 3: formaldehyde release by the flask method, European Standard, March 1996], perforator method [EN 120, 1993. Wood based panels—determination of formaldehyde content—extraction method called perforator method, European Standard, September 1993], Japanese test methods: desiccator methods [JIS A 1460, 2001. Building boards. Determination of formaldehyde emission—desiccator method, Japanese Industrial Standard, March 2001 and JAS MAFF 233, 2001] and small chamber method [JIS A 1901, 2003. Determination of the emission of volatile organic compounds and aldehydes for building products—small chamber method, Japanese Industrial Standard, January 2003], for solid wood, particleboard, plywood and medium density fiberboard. The variations between the results from different methods can partly be explained by differences in test conditions. Factors like edge sealing, conditioning of the sample before the test and test temperature have a large effect on the final emission result. The Japanese limit for F **** of 0.3 mg l1 (in desiccator) for particleboards was found to be equivalent to 0.04 mg m3 in the European chamber test and 2.8 mg per 100 g in the perforator test. The variations in inter-laboratory tests are much larger than in intra-laboratory tests; the coefficient of variation is 16% and 6.0% for the chamber method, 25% and 3.5% for the gas analysis method and 15% and 5.2% for the desiccator method. r 2006 Elsevier Ltd. All rights reserved. Keywords: Emission testing; Formaldehyde; Chamber; Building materials; Wood; Particleboard; Wood-based boards
1. Introduction
Corresponding author. Fax: +46 8 6428399.
E-mail address: [email protected] (M. Risholm-Sundman). 1352-2310/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2006.10.079
The emission of formaldehyde is an important factor in the evaluation of the environmental and health effects of wood-based board materials. Different test methods are used in different countries.
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
3194
In a global market it is of vital importance to be able to compare products with formaldehyde emission classes like E1 in Europe with F *** and F **** in Japan. It is also important that interlaboratory tests are performed to make it meaningful to compare results and possible to evaluate the variation that can be expected in the results obtained by different laboratories. In this article, comparisons of formaldehyde emissions measured with different test methods on solid wood, particleboard, plywood and medium density fiberboard (MDF) are presented. The impact of different test conditions, sample treatments, etc. are shown and discussed. 2. Standard test methods Determination of formaldehyde emission according to reference methods as the European chamber method EN717-1 requires, for example, chamber facilities and measurements until steady-state condition. There is, however, also a need for more
simple and less time-consuming standard test methods for production control and similar tasks. As some of the commonly used simple methods are performed at increased temperature (as EN717-2 and EN717-3) or includes extraction with solvents (as EN120), it cannot be expected that these methods always result in evaluations in compliance with results obtained by the emission chamber reference method. In Table 1, some main characteristics of the methods are given and they are also discussed in an article by Yu and Crump (1999). 2.1. Chamber methods The European chamber method, EN 717-1, is the reference method for the evaluation of the formaldehyde emission. The sample is placed in a chamber, normally 1 or 0.225 m3 in volume. The loading factor is 1 m2 m3 and the air exchange rate 1 h1. The temperature is held at 23 1C and the relative humidity (RH) at 45%. Formaldehyde released from the test pieces mixes with the air in
Table 1 Comparison of standard methods for the determination of formaldehyde emissions Method
Europe
Japan
Global a
EN 717-1 0.225, 1 or 412 m3 chamber EN 717-2 gas analysis 4 l chamber EN 717-3 500 mL flask EN 120 perforator JIS A 1901 20 l – 1 m3 chamber JIS A 1460 9–11 l desiccator JAS 233 9–11 l desiccator ISO/CD 12460 1 m3 chamber
Test sample
Test conditions
Size loading factor
Edge sealing (m open edge m2)
Temp/RH
Temp/RH
Air exchange/ hour
1 m2 m3
Partly (1.5 m m2)
23 1C/45%a
23 1C/45%
1
0.4 0.05 m
Yes
Not stated
60 1C/p 3%
15
0.025 0.025 m, 20 g 0.025 0.025 m, 110 g
No (80 m m2) No
Not stated
40 1C/~100%
No
Not stated
Toluene extraction at 110 1C
No
2.2 m2 m3
Yes
28 1C/50%
28 1C/50%
0.5
0.18 m2
No (27 m m2)
20 1C/65%
20 1C/0–80%b
No
0.18 m2
No (27 m m2)
Noc
1 m2 m3
Partly (1.5 m m2)
23 1C/50%a
Conditioning in the chamber, the values reported are steady-state values. See Fig. 4. c Stored at 20 1C for 1 day wrapped in plastic before testing. b
Conditioning
No 23 1C/50%
1
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
the chamber, and a specified volume of air is drawn from the chamber twice a day. The formaldehyde is absorbed in impinger flasks containing water and determined photometrically. The result of the test is given as the steady-state value in mg m3 or ppm (ml m3). In the Japanese 20 L chamber, JIS A 1901, the loading factor is normally 2.2 m2 m3 and the air exchange rate 0.5 h1. The temperature is 28 1C and the RH 50%. For this method, the result is given as the specific emission rate (EF) in mg m2 h1. When the loading factor and air exchange rate is 1, as in the European chamber, the concentration result in mg m3 is equal to the EF result in mg m2 h1. 2.2. Gas analysis The gas analysis method, EN 717-2, describes determination of accelerated formaldehyde release from wood-based panels. A test piece of known surface area is placed in a chamber with controlled temperature, RH, airflow and pressure. Formaldehyde released from the test piece mixes with the air in the chamber, and this air is drawn from the chamber. The formaldehyde is absorbed in water and determined photometrically. Particularly, the high temperature (60 1C), low RH (p3%) and high air exchange rate (15 h1) differ considerably from the chamber method and normal room conditions.
3195
based panels by extraction in a perforator. The formaldehyde is extracted from test pieces by means of boiling toluene and then transferred into water. The formaldehyde content of this aqueous solution is determined photometrically and expressed in weight per 100 g of dry board and is corrected for moisture content of 6.5%. In the standard it is noted that the correlation between perforator value and the formaldehyde emission depends on type of board material. Particularly, this difference between board materials in density, porosity and moisture can result in poor correlation between perforator values and emission results obtained in chamber methods. The perforator method is still a widespread method for production control. It is however often questioned from a work environment point of view due to use of toluene. 2.5. Desiccator method The two Japanese desiccator methods JIS A1460 and JAS MAFF 233 both describe determination of formaldehyde release from wood-based materials. Test pieces are placed in a desiccator containing a vessel with water. The formaldehyde released from the test pieces at 20 1C during 24 h is absorbed by the water and determined photometrically. As in the flask method, the RH is very high. 2.6. Comparison of different methods
2.3. Flask method The flask method, EN 717-3, describes determination of formaldehyde release from wood-based panels. Test pieces of known mass is suspended over water in a closed container at constant temperature. The formaldehyde released from the test pieces is absorbed by the water and determined photometrically. Particularly, the small test pieces with relatively large open edge area and large risk of sampling inhomogeneity, the container without ventilation, relatively high temperature (40 1C) and the high humidity generated in the container differ considerably from the chamber method and normal room conditions. 2.4. Perforator method The perforator method, EN 120, describes determination of formaldehyde content in wood-
There are large differences between European test methods as chamber method (EN 717-1), gas analysis method (EN 717-2), flask method (EN 717-3) and perforator method (EN 120) and Japanese test methods as desiccator (JIS A 1460; JAS 233) and the small chamber method (JIS A 1901). Test conditions like temperature, RH and air exchange rate vary substantially (see Table 1). Another important factor for differences in the emission results is the variation in required sample treatment, like (1) sealing of the edges, (2) sealing of the back and (3) conditioning before measurement. In the Japanese desiccator methods and the European flask method, the edges are not sealed even though they constitute a large part of the total area. Since the emissions in most cases are highest from the edges, this will have great impact on the results from these methods. Also considerable impact can be expected due to differences in conditioning; the gas analysis and JAS desiccator methods are run directly
ARTICLE IN PRESS 3196
M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
after removal of the plastic wrapping, while the other methods give values after conditioning of the samples in specified temperature and humidity. 3. Formaldehyde measurements included in this study 3.1. Inter-laboratory tests Samples of particleboard and MDF were sent to different laboratories for formaldehyde emission tests according to EN 717-1, EN 717-2 and JIS A 1460. The following comparisons were made: Chamber method, EN 717-1: Particleboards with thickness 10, 16 and 28 mm were used. Samples packed in plastic were sent to six laboratories in Europe and the tests were performed in both 225 l and 1 m3 chambers. The samples were cut at the laboratories and the edges partly sealed (1.5 m open edge m2). The results after 2 weeks in the chamber were compared (Winter Funch, 2004). Gas analysis method, EN 717-2: Particleboards with thickness 19 mm and MDF with thickness 10 and 22 mm, were used. Samples, cut to size and packed in plastic, were sent to 15 laboratories in Europe and Asia. The samples were analysed directly after opening the plastic and sealing the edges. Desiccator method, JIS A 1460: Two different particleboards, thickness 12 mm, and one MDF, thickness 12 mm were used. Samples were cut to size, packed in plastic and sent to eight laboratories in Europe and Asia. The samples were analysed after one week open conditioning at 20 1C, 65% RH. 3.2. Intra-laboratory tests For the determination of the reproducibility within one lab (tests made at different times), we made the following measurements for the different test methods: Chamber method, EN 717-1: Two MDF samples, which both gave constant concentrations in the chamber (mean values 0.19 and 0.15 mg m3, respectively), were measured 9 and 12 times respectively during a period of 14 days. A combined coefficient of variance (CV) was calculated. Gas analysis method, EN 717-2: Two calculations have been made. In the first, 12 test pieces were cut from a 12 mm MDF board and then wrapped in plastic. On six occasions during July 2004–January 2005, two of the pieces were run. A pooled standard deviation was calculated from these six duplicate results.
The second calculation was made from a total of 39 different samples (particleboards, MDF, plywood and edge glued panel samples) run during 2003. Also from these runs, a pooled standard deviation was calculated. Desiccator method, JIS A 1460: From a 12 mm MDF board 99 test pieces were cut and then stored open at 20 1C, 65% RH for 4 weeks. Nine pieces were put in each of eleven desiccators and the formaldehyde emission was measured according to the standard. After the measurement, the pieces were again stored open at 20 1C, 65% RH for 4 weeks and then measured again. This procedure was repeated four additional times and the CV for all measurements was calculated. 3.3. Comparison of different methods The formaldehyde emissions from solid wood, particleboards, plywood and MDF have been measured with different methods. The Japanese desiccator method JIS A 1460 was compared with European methods perforator EN 120 and chamber EN 717-1. Particleboards, thickness 12–19 mm, mostly of F **** quality were used, 21 samples for the EN 120 comparison and 8 for the EN 717-1 comparison. 3.4. Desiccator tests to determine RH in desiccators during test The RH was measured during the 24 h desiccator test using a capacitive humidity sensor, Testo 650, placed at the level of the sample. The humidity was measured in the desiccators as follows: 1. Reference, desiccator without sample. 2. Desiccator with an MDF sample conditioned according to JIS A 1460 at 20 1C, 65% RH for a week. The moisture content (MC) of the sample was 5.9% before conditioning and 8.2% after. 3. Desiccator with an interior plywood sample conditioned according to JAS MAFF 233 in plastic at 20 1C for 24 h. The moisture content of the sample was 7.6%. 3.5. Chamber tests to determine impact of temperature Four samples from the same MDF board were tested according to JIS A 1901, two at 23 1C and two at 28 1C in order to compare results from the
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
European and Japanese chamber tests. The emission was measured after both 7 and 14 days. 4. Results and discussion 4.1. Reproducibility in inter- and intra-laboratory tests The results from the inter-laboratory tests of methods EN 717-1, EN 717-2 and JIS A 1460 are given in Figs. 1–3. The CV, both within one laboratory and in the inter-laboratory tests, is presented in Table 2. Before the calculation of the CV for the results, each sample in the inter-laboratory tests were first normalized against the average of the results from all laboratories for the respective samples. As expected, the CV for the inter-laboratory tests is much higher than the intralaboratory CV. For the 39 board samples (of different kinds), the CV was calculated to 10%. This is much higher than the CV for the 12 pieces from the same sample, which is 3.5% and represents the spread in the method (excluding sample differences) (Fig. 4). 4.2. Comparison of different test methods The formaldehyde emission measured with different methods may be compared with the E1 limit of 0.1 ppm ( ¼ 0.124 mg m3) for chamber tests according to EN 717-1 and with the F **** and
3197
F *** limits of 0.3 and 0.5 mg l1 respectively for desiccator tests according to JIS A 1460 and JAS 233. If the results for the different samples are normalized against the values for particleboard E0, the diagram in Fig. 5 may be drawn. As can be seen, the other particleboard, E1, has approximately the same value (about 2) for each method. This shows that for the same kind of material, the methods show similar results; sample E1 emits twice as much as sample E0. Looking at the MDF sample, EN 717-1, EN 120 and JIS A 1460 show similar behaviour, this sample emits about three times as much as sample E0 according to all three methods. The gas analysis method, EN 717-2, on the other hand, gives a much more pronounced difference, while the flask method, EN 717-3, gives a value comparable to sample E1 (twice E0). The results for the plywood samples are similar for EN 717-1 and EN 717-2, while EN 717-3 gives a much higher value. This is probably at least partly explained by the huge difference in the ratio (open edge area)/(surface area) for the flask method compared to the other two. Solid wood gives low formaldehyde emission, close to the determination limit, for all methods. The result for the MDF sample, when tested in the European chamber according to EN 717-1 at 23 1C, was 0.10 mg m3. According to the conversion equation (ASTM D6007-96) this corresponds to 170 mg m3 at 28 1C. Since the loading factor and
Interlaboratory test, chamber method EN 717-1 10 mm PB, mean value 0.20 mg/m3
1.4
16 mm PB, mean value 0.18 mg/m3 28 mm PB, mean value 0.15 mg/m3
Relative response
1.2 1.0 0.8 0.6 0.4 0.2 0.0 1 - 1 m3
2 - 1 m3 3 - 1 m3 4 - 225 l 5- 225 l Laboratories - chamber size
6- 225 l
Fig. 1. Results from inter-laboratory tests performed according to EN 717-1 with MDF and particleboards. The results for each test have been normalized against the average of the results from all laboratories for the respective sample.
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
3198
Interlaboratory test, gas analysis method EN 717-2 19 mm PB, mean value 8.3 mg/m2,h
1.8
10 mm MDF, mean value 3.0 mg/m2,h 22 mm MDF, mean value 2.9 mg/m2,h
Relative response
1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 1
2
3
4
5
6
7 8 9 Laboratories
10
11
12
13
14
15
Fig. 2. Results from inter-laboratory tests performed according to EN 717-2 with MDF and particleboards. The results for each test have been normalized against the average of the results from all laboratories for the respective sample.
Interlaboratory test, desiccator method JIS A 1460 12 mm PB, mean value 0.24 mg/L 12 mm PB, mean value 0.64 mg/L 12 mm MDF, mean value 0.59 mg/L
1.40
Relative response
1.20 1.00 0.80 0.60 0.40 0.20 0.00 1
2
3
4 5 Laboratories
6
7
8
Fig. 3. Results from inter-laboratory tests performed according to JIS A 1460 with MDF and particleboards. The results for each test have been normalized against the average of the results from all laboratories for the respective sample.
the air exchange rate both are 1, this gives an emission rate of 170 mg m2 h1. The explanation for the substantially higher value obtained in the European chamber compared to the result in the Japanese chamber, 74 mg m2 h1, must be that the sample is run with partly open edges (1.5 m m2 open edge). It has been shown earlier (on parquet samples), that the emission from the edges is much higher than from the surface (Risholm-Sundman and Wallin, 1999).
The effect of conditioning may be seen in Table 3 from the results on solid pine obtained with the desiccator tests. In JAS 233, when the sample is wrapped in plastic until the test starts, the emission was measured to 0.18 mg l1. The conditioning for 1 week in JIS A 1460 at 20 1C 65% RH lowered the formaldehyde emission dramatically, to o0.1 (0.05) mg l1. The difference between our result in the gas analysis of solid pine, 0.3 mg m2 h1 and the value obtained by Meyer and Boehme (1997),
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
3199
Table 2 Reproducibility in inter- and intra-laboratory tests Method
Intra-lab No. of tests
EN 717-1 .225, 1 or 12 m3 chamber 21 EN 717-2 gas analysis 4 l chamber 12a 39b JIS A 1460 9–11 l desiccator 66
Inter-lab CV (%)
No. of labs
No. of samples CV (%) per lab
Outliers excluded
CV (%)
6.0 3.5a 10b 5.2
6 15
3 3
16 25
c
11 17
8
3
15
d
a
Test pieces cut from the same sample. Duplicate analyses of test pieces from 39 different board samples. c Result of one sample from one of six laboratories excluded (see Fig. 1). d All results from two of 15 laboratories excluded (see Fig. 2). b
100 90 80
RH (%)
70 60 50 40 JIS A 1460, MDF MC=8.2% JAS 233, Plywood MC=7.6% Empty desiccator
30 20 10 0 0
5
10
15
20
Hours Fig. 4. The RH in desiccators with samples of different moisture content (MC), as a function of time.
8 7 Normalised value
6 5
Value:16
EN 717-1 EN 717-2 EN 717-3 EN 120 JIS A 1460
4 3 2 1 0 Solid oak
Solid pine
PB E0
PB E1
Plywood int
Plywood ext
MDF
Fig. 5. Comparison of the results from different test methods on different samples. The results have been normalized against the particleboard (PB) E0 value.
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
3200
Table 3 Formaldehyde emission measured according to different standard methods Material
Solid wood
Oak MC ¼ 8% Pine MC ¼ 8%
Thickness
Chamber
Gas analysis
Flask
Perforator
Chamber
Desiccator
(mm)
EN 717-1 (mg m3)
EN 717-2 (mg m2 h1)
EN 717-3 (mg kg1)
EN 120 (mg per 100 g)
JIS A 1901 (mg m2 h1)
JIS A 1460 (mg l1)
JAS 233 (mg l1)
20
0.005a
0.05a
0.06a
0.19a
o0.1 (0.00)c
o0.1 (0.03)c
20
0.006a
0.3b
0.16a
0.23a
o0.1 (0.05)c
0.18
a
Particle board
Plywood
MDF
E0 MC ¼ 6.8% E1 MC ¼ 6.0% Interior MC ¼ 8.6% Exterior MC ¼ 6.2% MC ¼ 5.9%
12
0.03
0.09 0.8
2
2–3
0.2
12
0.07
2
4
4.6
0.5
0.20
5.5
15
0.01
0.2
1.4
12
0.10
4.4
3.5
9.5
32
4.7 0.3 7.5
74 (43 at 23 1C)
0.6
a
Value from Meyer and Boehme (1997). No conditioning. c Analysed value, below the determination limit of 0.1 mg l1. b
0.9
mg L-1 y = 6.8561x + 0.0463
0.8
R2 = 0.7294
JIS A 1460
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
EN 717-1
0.09
0.1
mg m-3
Fig. 6. Correlation between JIS A 1460 and EN 717-1 for particleboards, thickness 12–19 mm.
0.09 mg m2 h1 may also be an effect of different conditioning. The correlation between the Japanese desiccator method JIS A 1460 and the European methods EN 717-1 (chamber) and EN 120 (perforator) was not convincing, the R2 values obtained were 0.66 and 0.73 (see Figs. 6 and 7). The values for the European methods that correspond to F ****, 0.3 mg l1, are given in Table 4. These values agree quite well with values reported earlier (Bulian et al., 2004) for particleboards in the range of 0.01–0.1 mg m3.
4.3. Desiccator tests—RH The results from the humidity measurements in the desiccators show that wood-based samples will lower the RH from 100% to about 60–80% (see Fig. 4). 4.4. Chamber tests—temperature The MDF sample was tested according to JIS A 1901 at both 23 and 28 1C. The result obtained at 23 1C,
ARTICLE IN PRESS M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
3201
mg L-1
JIS A 1460
0.9 0.8
y = 0.0825x + 0.0816
0.7
R2 = 0.7559
0.6 0.5 0.4 0.3 0.2 0.1 0 0
1
2
4
3
5
6
7
Perforator, EN 120
8 mg / 100 g
Fig. 7. Correlation between JIS A 1460 and EN 120 for particleboards, thickness 12–19 mm.
Table 4 Values in chamber test EN 717-1 and perforator test EN 120 corresponding to the F **** limit, 0.3 mg l1, in JIS A 1460
EN 717-1 (mg m3) EN 120 (mg per 100 g)
Result corresponding to F **** limit
Literature result (Bulian et al., 2004)
0.037 2.8
0.04 2.4
43.0 mg m2 h1, may be converted to 28 1C by using the conversion equation, C ¼ C0 e9799(1/t1/t0), according to ASTM D6007-96. This gives a value of 75, which agrees very well with the measured value 74, given in Table 3. 5. Conclusion Comparisons of different formaldehyde emission methods show that: The variations can be explained by differences in the test conditions. Factors like edge sealing, conditioning of the sample before the test and test temperature have a large effect on the final emission result. The Japanese limit for F **** of 0.3 mg l1 (in desiccator) for particleboards was found to be equivalent to 0.04 mg m3 in the European chamber test and 2.8 mg per 100 g in the perforator test. The correlation between the desiccator JIS A 1460 and the chamber and perforator methods respectively is, however, not convincing ðR2 ¼ 0:7Þ. The variations in inter-laboratory tests with chamber method EN 717-1 (CV ¼ 16%), gas
analysis EN 717-2 (CV ¼ 25%) and desiccator JIS A 1460 (CV ¼ 15%) are much higher than in intralaboratory tests. Different standards in different countries make it difficult for the manufacturer of wood-based board materials to be able to comply with regulations in different countries. There is a need for international harmonization. Standardization work within ISO should be encouraged. References ASTM D6007-02, 2002. Standard test method for determining formaldehyde concentration in air from wood products using small-scale chamber. American Standard, April 2002. Bulian et al., 2004. Formaldehyde testing of wood-based panels: correlations between European and Japanese test methods. In: Wood-based Panel Symposium in Hannover, September 2004. EN 120, 1993. Wood-based panels—determination of formaldehyde content—extraction method called perforator method. European Standard, September 1993. EN 717-1, 2004. Wood-based panels—determination of formaldehyde release—Part 1: formaldehyde emission by the chamber method. European Standard, October 2004.
ARTICLE IN PRESS 3202
M. Risholm-Sundman et al. / Atmospheric Environment 41 (2007) 3193–3202
EN 717-2, 1994. Wood-based panels—determination of formaldehyde release—Part 2: formaldehyde release by the gas analysis method. European Standard, November 1994. EN 717-3, 1996. Wood-based panels—determination of formaldehyde release—Part 3: formaldehyde release by the flask method. European Standard, March 1996. ISO/DIS 12460, 2005. Wood-based panels—determination of formaldehyde release—formaldehyde emission by the 1 m3 chamber method. Draft International Standard, January 2005. JAS 233, 2003. Japanese Agricultural Standard for plywood, Japanese Agriculture Standard, February 2003. JIS A 1901, 2003. Determination of the emission of volatile organic compounds and aldehydes for building products—small chamber method. Japanese Industrial Standard, January 2003.
JIS A 1460, 2001. Building boards. Determination of formaldehyde emission—desiccator method. Japanese Industrial Standard, March 2001. Meyer, B., Boehme, C., 1997. Formaldehyde emission from solid wood. Forest Products Journal 47 (5), 45–48. Risholm-Sundman, M., Wallin, N., 1999. Comparison of different laboratory methods for determining the formaldehyde emission from three-layer parquet floors. Holz als Rohund Werkstoff 57, 319–324. Winter Funch, L., 2004. Danish Technological Institute. Report RRT on prEN 717-1, 2004.08.25. Yu, C.W.F., Crump, D.R., 1999. Testing for formaldehyde emission from wood-based products—a review. Indoor Built Environment 8, 280–286.