A novel assessment of odor sources using instrumental analysis combined with resident monitoring records for an industrial area in Korea

Atmospheric Environment 74 (2013) 277e290

Contents lists available at SciVerse ScienceDirect

Atmospheric Environment journal homepage: www.elsevier.com/locate/atmosenv

A novel assessment of odor sources using instrumental analysis combined with resident monitoring records for an industrial area in Korea Hyung-Don Lee a, Soo-Bin Jeon a, Won-Joon Choi b, Sang-Sup Lee c, Min-Ho Lee d, Kwang-Joong Oh a, * a

Department of Environmental Engineering, Pusan National University, San 30 Jangjeon-Dong, Busan 609-735, Republic of Korea Fuel Cell System Development Team, Doosan Heavy Industries & Construction, 463-1 Jeonmin-Dong, Daejeon 305-811, Republic of Korea Department of Environmental Engineering, Chungbuk National University, 52 Naesudong-ro, Cheongju 361-763, Republic of Korea d Environment & FireTeam-Ulsan Plant, Hyundai Motor Company, 700 Yangjung-Dong, Ulsan 683-791, Republic of Korea b c

h i g h l i g h t s
We make an attempt novel assessment for characteristics of odorants in industrial area.
The concentration of RSCs was significantly higher than other odorous compounds.
The offensive odors in residential area were characterized as ‘burned’ and ‘other’ smells.
We confirm a strong correlation between instrumental analysis and resident monitoring data.
Resident monitoring data can be used effectively to evaluate the characteristic of odorants emitted in industrial area.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 20 November 2012 Received in revised form 28 March 2013 Accepted 2 April 2013

The residents living nearby the Sa-sang industrial area (SSIA) continuously were damaged by odorous pollution since 1990s. We determined the concentrations of reduced sulfur compounds (RSCs) [hydrogen sulfide (H2S), methyl mercaptan (CH3SH), dimethyl sulfide (DMS), and dimethyl disulfide (DMDS)], nitrogenous compounds (NCs) [ammonia (NH3) and trimethylamine (TMA)], and carbonyl compounds (CCs) [acetaldehyde and butyraldehyde] by instrumental analysis in the SSIA in Busan, Korea from Jun to Nov, 2011. We determined odor intensity (OI) based on the concentrations of the odorants and resident monitoring records (RMR). The mean concentration of H2S was 10-times higher than NCs, CCs and the other RSC. The contribution from RSCs to the OI was over 50% at all sites excluding the A-5 (chemical production) site. In particular, A-4 (food production) site showed more than 8-times higher the sum of odor activity value (SOAV) than the other sites. This suggested that the A-4 site was the most malodorous area in the SSIA. From the RMR analysis, the annoyance degree (OI  2) was 51.9% in the industrial area. The ‘Rotten’ smell arising from the RSCs showed the highest frequency (25.3%) while ‘Burned’ and ‘Other’ were more frequent than ‘Rotten’ in the residential area. The correlation between odor index calculated by instrumental analysis and OI from the RMR was analyzed. The Pearson correlation coefficient (r) of the SOAV was the highest at 0.720 (P < 0.05), and overall results of coefficient showed a moderately high correlation distribution range (from 0.465 to 0.720). Therefore, the overall results of this research confirm that H2S emitted from A-4 site including food production causes significant annoyance in the SSIA. We also confirm RMR data can be used effectively to evaluate the characteristic of odorants emitted from the SSIA. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Reduced sulfur compounds Resident monitoring records Odor intensity Odor activity value Pearson correlation coefficient

1. Introduction

* Corresponding author. Tel.: þ82 51 510 2417; fax: þ82 51 583 0559. E-mail address: [email protected] (K.-J. Oh). 1352-2310/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.atmosenv.2013.04.001

As a result of recent industrial development, nearby residents are directly and indirectly exposed to air pollutants in various forms in the environment. Especially, air pollution episodes caused

278

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290

by odorants are a particularly important pollution phenomenon in the atmospheric environment of a metropolis (Bundy, 1992). Odorants cause serious problems when an industrial complex is located close to a residential area, and some of the odorants emitted from such an industrial area can cause damage to the residential area. In addition, odorants can even be perceived at low threshold levels, and they can have a psychological impact on residents even at very low concentrations (Capelli et al., 2011; Mackie et al., 1998; Kabir and Kim, 2010; Herr et al., 2003). Odor can cause headaches, loss of appetite, gastrointestinal disorders, sleep disturbances, dyspneas, and allergic phenomena, and civil complaints of odors occur frequently (Gostelow et al., 2001; Tsao et al., 2011). Therefore, odorants emitted from industrial areas can lead to a lower overall quality of life and aggravate the health of the residents and have a negative effect on the local economy (Blanes-Vidal et al., 2012). The reduced sulfur compounds (RSCs) including hydrogen sulfide (H2S), methyl mercaptan (CH3SH), dimethyl sulfide (DMS), and dimethyl disulfide (DMDS) are identified as the major constituents of malodor problems. The government has designated odor management areas to reduce civil complaints of odorants, and the actual odorant emissions in areas where odor concentration is seriously high are periodically investigated (Yu et al., 2009). The Sa-sang industrial area (SSIA) was completed in 1975, and it is difficult to conduct environmental monitoring in the area because residential, commercial, and industrial zones are so concentrated. In particular, the environmental problem due to the underdeveloped manufacturing industry occupying most of the SSIA is serious. Thus, the Busan government consistently has tried to improve environmental quality, but the incidence rate of odor complaints from various causes has increased every year. An enormous residential area is located in the east of the SSIA. Odorants can easily move to the residential area owing to the westerly winds that prevail in the country. For this reason, there are constant odor complaints registered in the SSIA. This phenomenon increased markedly when residents moved in on a large scale in the early 1990s. Various aspects must be considered to analyze the cause and characteristics of odorants for the minimization of odor complaints (Emerson and Rajagopal, 2004; Kim et al., 2005). We have to accurately analyze the characteristics of the change in odorants throughout the year because there are numerous facilities that produce odor emissions located close together in the SSIA. Social participation and strong community involvement in detecting odor problems are needed to identify and evaluate odor sources causing irritation in areas surrounding the industrial area (Nicolas et al., 2010). There have been many recent studies employing social participation and questionnaires to evaluate the annoyance caused by odors (Blanes-Vidal et al., 2012; Sucker et al., 2008; Gallego et al., 2008; Susaya et al., 2011a; Heaney et al., 2011; Aatamila et al., 2012). The objective of this study is to quantify odorants emitted from the SSIA and to identify emissions sources and major odorants in the SSIA. Odorants such as RSCs, nitrogenous compounds (NCs), and carbonyl compounds (CCs) were sampled in the SSIA from June to November 2011, and their concentrations were determined using a gas chromatography (GC). The results were compared with emission standards and odor threshold values (OTVs). While odor activity value (OAV) and odor intensity (OI) were determined based on the odorants concentrations in many other studies (Kim and Park, 2008; Susaya et al., 2011a; Trabue et al., 2011; Seo et al., 2011; Parker et al., 2013), both the odorants concentrations and resident monitoring records (RMR) were used in this study. The concentrations of the odorants determined by the instrumental analysis were also correlated to the OI.

2. Materials and methods 2.1. Sampling sites Fig. 1 shows a map of the research area. The SSIA is close to a basin and is surrounded by mountains. Measurements were taken from 6 plant sites (A-1eA-6) and from residential areas (B-1eB-3) in the SSIA in Busan, Korea. Three residential areas (B-1eB-3) where odor complaints have been registered are located in the eastern part of the SSIA. Table 1 shows the types of major industry (A-1eA-6), including iron (metal) production, chemical production, food (fish, meat, cakes, and beverages) production, and waste and sewage treatment. Sampling was conducted during summer (June Aug) and autumn (OcteNov) in 2011. Measurements were performed daily in the morning (9e11 a.m.), afternoon (2e5 p.m.), and evenings (7e10 p.m.), 20 times each month and a total of 100 times. Table 2 shows the meteorological conditions (temperature, humidity, and wind conditions) at the research area. The wind direction was primarily northwesterly (WNW, NW, and NNW). 2.2. Instrumental operation The offensive odor prevention Law was implemented by the Korean Ministry of environment in 2004 in order to manage odors efficiently, and 22 substances have been designated as representative offensive odorants (KMOE, 2007a,b; Seo et al., 2011). In this study, among the representative offensive odorants measured were NCs including ammonia (NH3) and trimethylamine (TMA), which is known to cause a fishy smell at very low concentrations (Mushiroda et al., 1999; Zhang et al., 2005; Rappert and Muller, 2005), and RSCs including H2S, CH3SH, DMS, and DMDS. The CCs including acetaldehyde and butyraldehyde were also monitored. For NH3, sodium nitroprusside and sodium hypochlorite were added to the samples and analyzed at 640 nm (range: 190e 1100 nm) using a UV/vis spectrophotometer (Shimadzu, UV-1700, Japan) based on the indophenols method (Seo et al., 2011). The aldehyde composition collected from the DNPH cartridge was analyzed by high performance liquid chromatography (HPLC) (Varian, Pro Star 210, USA). Samples were eluted with 70% acetonitrile at 1.0 mL min1 and the converted concentration was measured using 20-ml samples with HPLC. A non-polar octadecyl silane (ODS) column was used for effective detection of different fractions and had an inside diameter of 4.6 mm and a length of 250 mm. RSCs samples were collected using tedlar bags (10 L, TDC, Japan) at 0.5 L min1 using vacuum suction equipment over 10 min after nitrogen gas of high purity (99.99%) had been used to clean the sampling bag five times. The samples were analyzed by GC/PFPD (Varian, GC CP-3800, USA). The column used for separation of the sample was a CP-SIL 5CB LOW Bleed/MS (60 m  0.25 mm, 1.0 mm). The quantification methods used a cold trap and chromatography/capillary column at 150  C and the flux of the column was regulated to 2 mL min1 and the temperature of the detector was at 250  C (Micone and Guy, 2007). The TMA was analyzed using GC-NPD (Shimadzu, GC-14B, Japan) with a column whose flux was set to 2 mL min1 and at 30  C for 15 min for adsorption and 260  C for 3 min for desorption (Chien et al., 2000). Table 3 shows the relative standard error (RSE) and the minimum detection limit (MDL) for quality assurance (QA) of the analytical chemistry. With respect to instrumental precision, the RSE values were calculated based on five repeated analyses with r2 > 0.989 and the values were 0.25 (TMA) to 5.40 (NH3), and the MDL values described in Table 3 and all of the MDL values of compounds were lower than the emission standard value in Korea.

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290

279

Fig. 1. Geographical location of the research area in Busan city, Korea.

2.3. RMR method Odorant monitoring data were collected from 200 participants. They consisted of 60% male and 40% female. 50% have lived in the SSIA neighborhood, and the other 50% have lived more than 5 km Table 1 The types of major business at the sampling sites in SSIA. Sampling sites

Type of industry

Detail

N

A-1

Iron production

20

A-2

Iron production

A-3

Chemical production

A-4

Food and beverage production Chemical production

Metal-working, nonmetal, plating, smelting, casting Metal processing, assembly metal, plating, smelting, melting Rubber, leather, bag, plastic, paint, shoes Food (starch, fish cake) feed, fodder, fertilizer, marine products Synthetic resin and detergent, paint and varnish, rubber, chemicals Wastewater and sewage, waste

A-5 A-6

Waste and sewage treatment

11 10 12 13 8

from the SSIA. SSIA residents did not participate here because local residents generally can judge odor sources more sensitively than other residents (Nicolas et al., 2010). The participants by age were: 20 years, 25%; 30 years, 25%; 40 years, 25%; and 50 years, 25%. Monitoring was conducted 4 times a week for JuneeAugust and OctobereNovember 2011. There were 3 sets of monitoring periods: morning (7e11 a.m.), afternoon (1e5 p.m.), and evening (7e10 p.m.) and the OI was estimated on a 6-level scale (0: none odor, 1: slight, 2: noticeable, 3: strong, 4: very strong, 5: unbearable) (Nicolas et al., 2010; Gallego et al., 2008) by direct olfactory observation. The following information was additionally recorded: (a) date and time; (b) weather data including precipitation, wind direction, wind speed, and temperature; and (c) other comments (Nimmermark et al., 2005). 200 participants were divided to 50 groups. Each group included 1 participant from each age period. A total of 33,600 questionnaires were gathered, except on rainy days, and odor observations were recorded together with a professional odor panelist. Monitoring was conducted at the same sites as the measurements (A-1eA-6 and B-1eB-3). Various odor samples were tested for 2 months to judge the OI of odorants before monitoring

280

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290

Table 2 Meteorological conditions at the sampling sites for each month. Temp ( C)

Relative humidity (%)

Range (mean)

Range (mean)

Range (mean)

June

15.2e27.8 (21.3)

35.3e85.7 (78)

0.5e12.1 (3.3)

July

19.4e33.0 (25.1)

37.4e89.9 (79)

0.0e13.6 (3.4)

August

19.5e32.3 (25.8)

43.0e86.2 (78)

0.0e11.1 (3.4)

October

6.4e27.4 (17.6)

21.2e64.1 (58)

0.6e10.0 (2.7)

November

0.9e24.6 (14.1)

10.2e70.3 (60)

0.5e9.8 (2.7)

Month

started. The questionnaire recorded the characteristic of the smell, wind direction, OI, and the other characteristics at each measurement site. The participants received remuneration for taking the survey sincerely and responsibly. Before starting field inspections,

Wind direction (wind rose)

Wind speed (m s1)

the participants were familiarized with the locality (Sucker et al., 2008). The OI values determined based on the RMR were analyzed using the SPSS program (version 18.0, USA). To analyze the OI

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290

281

Table 3 List of basic parameters of the analyzed compounds. Type

Nitrogenous compounds Reduced sulfur compounds

Carbonyl compounds a b c d

Odor descriptiona

Compounds

Ammonia Trimethylamine Hydrogen sulfide Methyl mercaptan DMS DMDS Acetaldehyde Butyraldehyde

Emission standard(ppb)b

Pungent Fishy Rotten egg Rotten onion, decayed cabbage Pungent and fruit Pungent and burned, sweaty

Industrial area

Other area

2000 20 60 4 50 30 40 5

1000 5 20 2 10 9 10 1

OTVc (ppb)

MDLd (ppb)

Precision (RSE in %)

r2

100 0.1 0.5 0.1 0.1 0.3 2 0.3

3.5 1.0E-02 1.7E-02 2.3E-03 3.5E-03 2.5E-03 1.33E-03 3.0E-03

5.40 0.25 0.67 0.80 0.33 0.54 1.68 1.26

0.990 0.992 0.991 0.991 0.990 0.989 0.993 0.989

Source: Rappert and Muller, 2005; Rosenfeld et al., 2001; Gostelow et al., 2001. Source of odorant emission guideline: KMOE (2007b). Source: KMOE (2007b). MDL: method detection limit.

confidence interval (CI), a P value of 0.05 was taken as the significance level and 95% confidence intervals were calculated (Sucker et al., 2008). We also assessed Cronbach’s alpha index to assess the reliability of RMR. 2.4. Analysis of odor indices Generally, people sense odorant concentrations differently because odorants have distinct OTV. Even if a high concentration of odorants occurs, it is not certain that a high OI will be sensed by olfactory observation. Therefore, the relative OI must be determined for each odorant. In this study, we used the OAV which divides the odor concentration by OTV and evaluates the sum of odor activity value (SOAV) (Feilberg et al., 2010; Trabue et al., 2006, 2011; Parker et al., 2013). The report by Kim and Park (2008), SOAV was referred to an identical meaning as the sum of odor quotient (SOQ). The contribution of a particular odorant was evaluated by using the SOAV and OAV.

Odor Activity ValueðOAVÞ ¼ SOAV ¼

X

Odor concentrationðppbÞ OTVðppbÞ

Odor Activity Value ¼

Odor ContributionðOCÞ ¼

X

OAV  100 SOAV

OAV

(1)

(2)

(3)

The human nose can detect and discriminate odors at concentrations even lower than those detectable by gas chromatography (Rappert and Muller, 2005). However, we limited the evaluation only to odor pollution degrees as the concentration of odorants. Generally, OI has been applied to determine the impact of odors in the ambient air (Jiang et al., 2006). The OI is one of the most important sensory properties of odorants. It is defined as the quantitative attribute of the stimulus and roughly correlated with the strength of the stimulus (Chen et al., 1999). The functional formula proposed by Nagata (Nagata, 2003) was used to convert concentration data to OI. Table 4 shows the functional formula for each odorant (Nagata, 2003) and the examples of the odor concentration (X) and the OI (Y) calculated from the formulae. The application of these formulae allows the conversion of concentrations to OI on a scale of 1 through 5. Although the formula was proposed by Japanese researcher, many Korean researchers (Choi et al., 2005; Kim et al., 2009; Kabir and Kim, 2010; Kim, 2011) have used the formula because Korean environmental regulations are similar to Japanese environmental regulations regarding the OTVs of the odorants (Kim et al., 2009). The OI value obtained using the Nagata formula was termed as computed odor

intensity (COI) while the value obtained from RMR was termed as OI in this paper. We performed a Pearson correlation between odorant concentration by instrument analysis and direct olfaction by RMR at the SSIA. Because it is difficult to directly correlate the odorant concentration with OI, we employed several indices (SOC: the sum of odor concentration, SCOI: the sum of computed odor intensity, and SOAV). The SCOI values were calculated by combining the COI values of an individual odorant on a logarithmic scale, and the SCOI equation (Kim and Park, 2008; Kim, 2011) is:

  SCOIi ¼ log 10COIðiÞ1 þ 10COIðiÞ2 þ 10COIðiÞ3 þ /// þ 10COIðiÞ8

(4) where, COI(i) is the odor intensity of each odorant. 3. Results and discussion 3.1. Concentration of odorants at the industrial and residential sites A direct evaluation of concentrations of different odorants is meaningful for the understanding of fundamental aspects of odorant emissions from strong sources (Kim et al., 2006). Table 5 shows the concentration (ppb) of odorants measured at all sampling sites, and all average concentration of odorants measured for the entire study period (JuneNov) are presented in Fig. 2. The mean concentrations of all the RSCs (H2S: 36.50 ppb, CH3SH: 2.28 ppb, DMS: 2.23, and DMDS: 1.61 ppb), NCs (NH3: 760. 50 ppb and TMA: 2.67 ppb) and CCs (acetaldehyde: 11.81 and butyraldehyde: 2.67 ppb) were highest at the A-4 site. Of the RSCs, the concentration of H2S was the highest at 103.43 ppb (40.33  58.02 ppb) in autumn. Of the NCs, the maximum concentration was for NH3 at 1279.15 ppb (777.37  586.47 ppb) in summer, and of the CCs, acetaldehyde was highest at 32.87 ppb (14.23  19.86 ppb) in autumn (Table 5). The total mean concentrations of the odorants during the measurement period were in the order NH3 (428.39 ppb), H2S (10.46 ppb), acetaldehyde (3.78 ppb), butyraldehyde (1.52 ppb), DMS (0.97 ppb), DMDS (0.90 ppb), TMA (0.85 ppb), CH3SH (0.76 ppb) at the SSIA (A1eA-6), and NH3 (54.16 ppb), H2S (0.54 ppb), acetaldehyde (0.36 ppb), butyraldehyde (0.10 ppb), DMS (0.06 ppb), CH3SH (0.04 ppb), DMDS (0.03 ppb), and TMA (0.01 ppb) in the residential areas (B-1eB-3). Generally, the mean concentration of NH3 was higher than the concentrations of the other odorants. The concentration of H2S was higher than the other RSCs and varied over a large range (0.05e103.43 ppb), in agreement with previous research (Seo et al., 2011; Kim et al., 2005; Susaya et al., 2011a; Lee et al., 2006). Although the mean concentration of H2S was lower

0.0015 0.015 0.047 0.14 0.45 1.4 14 0.00011 0.0014 0.0052 0.019 0.067 0.24 3.1 0.15 0.59 1.2 2.4 4.7 9.3 37 a

b

0.00012 0.00065 0.0016 0.0041 0.01 0.026 0.16 0.0005 0.0056 0.019 0.063 0.21 0.71 8 1 2 2.5 3 3.5 4 5

Nagata (2003). Formula: computed odor intensity (Y) and odor concentration (X: ppm).

0.00028 0.0029 0.0092 0.03 0.094 0.3 3.1

DMDS DMS CH3SH H2S

Equivalent of odor concentration (X: ppm) COI value (Y)

0.00012 0.0023 0.01 0.0445 0.19 0.84 1.6

Butyraldehyde Acetaldehyde NH3

TMA

Nitrogenous compounds (NCs) NH3 TMA Y ¼ 1.670log X þ 2.38 Y ¼ 0.901log X þ 4.56 DMDS Y ¼ 0.985log X þ 4.51 DMS Y ¼ 0.784log X þ 4.06 Type Reduced sulfur compounds (RSCs) Material of odorant H2S CH3SH Functional formulaa Y ¼ 0.950log X þ 4.14b Y ¼ 1.250log X þ 5.99 Example of X and Y

Table 4 Functional formula for conversion of concentration to computed odor intensity (COI) for odorous compounds.

0.00032 0.0029 0.0089 0.027 0.084 0.26 2.3

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290

Carbonyl compounds (CCs) Acetaldehyde Butyraldehyde Y ¼ 1.010log X þ 3.85 Y ¼ 1.030log X þ 4.61

282

than the emissions standard (KMOE, 2007b), its maximum concentration exceeded the emissions standard (Table 2) and was 207 times higher than the OTV (0.5 ppb). The mean concentration of H2S exceeded the OTV by a factor of 21 and was approximately 10 times higher than other RSCs at all sampling sites. Although the maximum concentration (1279.15 ppb) of NH3 was higher than that of other odorants, the OTV was higher than other odorants by a factor of more than 100. Hence, OI and the annoyance registered by residents were lower than that of H2S. The maximum concentration of DMS, CH3SH, and TMA exceeded the OTV (0.1 ppb) by factors of 40 and 43, and butyraldehyde exceeded the OTV (0.3 ppb) by a factor of 32.9, so OI value detected by residents was actually higher than that of NH3. Hence, the concentration distribution of H2S, CH3SH, and DMS among the RSCs, TMA among the NCs, and butyraldehyde among the CCs caused significant malodor phenomena at this study area, as noted in previous research (Choi et al., 2005). We identified that the quantitative contributions of H2S to the emission levels of RSCs was extremely high in various emission sources in the SSIA. Most odorous compounds emitted from industrial plants are not a public health concern but can be considered a public nuisance (Rappert and Muller, 2005). If such emission levels are continuously maintained, residents who work or live in the area are exposed to health and psychological risks such as stress, feeling of aversion, displeasure, anxiety, and annoyance. Contrary to the report by Susaya et al. (2011a), the seasonal variations of the mean concentration of H2S and CH3SH were slightly higher during autumn, and DMS and DMDS showed similar trends. They reported a nighttime mean for H2S higher than its daytime mean and an increased relative humidity and dew point during the nighttime because of the shorter daytime in autumn. We have analyzed this result, and the concentration in summer was lower than the concentration in autumn as atmospheric conditions (wind direction/speed, temperature) during the measurement period. The geographical characteristic of the research area was distinctly different compared to other industrial areas. Additionally, the concentration of CH3SH during autumn was higher than in summer because the rate of photodecomposition of CH3SH was very high in the daytime (Toda et al., 2010), and the atmospheric lifetime was very short from a few minutes to hours (Smet et al., 1998). DMS and DMDS were formed by methylation reactions between amino acids by decomposition processes of proteins and H2S (Lomans et al., 2002). Hence, we identified a high concentration distribution of H2S at the sampling sites (A-4) where there were also high mean concentrations of DMS (2.37 ppb) and DMDS (1.70 ppb). In case of CCs, the mean concentration of acetaldehyde produced by feed and food-making processes (Rappert and Muller, 2005) exceeded the OTV (2 ppb) by a factor of 5.9. Fermentation and dry food processing emit many odorous materials such as acetaldehyde. TMA produced by microbial degradation of fish constituents (Seo et al., 2011) had a high mean concentration (2.67 ppb) at the A-4 site. This was generally attributed to microbial degradation, which was expedited by high temperature and relative humidity in the summer. Although the results provided quantitative information for each odorant, the overall strengths of each odorant could not be assessed by simply comparing their concentration distributions (Kim et al., 2009). Therefore, we assessed the concentration data for offensive odor components by converting them to OAV and OI as an indirect means of comparing the relative strengths of different industry types. 3.2. Evaluation of OAV and OI It was important to analyze the concentration of odorants because it helped us elucidate the fundamental distribution of

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290

283

Table 5 An overall summary of odorant concentrations (ppb) measured at sampling sites in this study. Type Industrial areas A-1 site RSCs (N ¼ 82)

NCs (N ¼ 86) CCs (N ¼ 85) A-2 site RSCs

NCs CCs A-3 site RSCs

NCs CCs A-4 site RSCs

NCs CCs A-5 site RSCs

NCs CCs A-6 site RSCs

NCs CCs Resident areas B-1 site RSCs

NCs CCs B-2 site RSCs

NCs CCs

Max

Mina

Compound

Mean

SD

H2S CH3SH DMS DMDS NH3 TMA Acetaldehyde Butyraldehyde

2.08/2.57b 0.21/0.36 0.52/0.44 0.38/0.44 385.00/220.33 0.17/0.02 1.50/1.75 0.63/1.28

2.13/3.37 2.11/0.13 0.32/0.52 0.11/0.39 207.48/99.08 1.02/0.03 1.56/0.71 1.28/0.52

7.32/9.11 0.30/0.51 0.59/1.03 0.72/0.64 580.00/308.27 0.62/0.22 2.39/3.99 2.18/1.89

0.09/0.05 0.02/0.02 0.02/0.01 0.01/0.01 12.40/35.60 0.01/0.01 0.13/0.05 0.02/0.07

H2S CH3SH DMS DMDS NH3 TMA Acetaldehyde Butyraldehyde

4.98/2.10 0.06/0.16 0.15/0.15 0.66/0.51 175.38/210.30 0.03/0.04 2.00/1.63 1.06/3.46

3.25/2.53 1.03/0.04 0.15/0.21 0.58/0.24 160.99/107.36 0.01/0.02 2.64/1.09 0.55/5.99

15.26/12.90 0.26/0.20 0.33/0.37 0.98/1.01 270.32/345.51 0.38/0.18 3.87/3.72 1.79/9.87

0.27/0.41 0.03/0.01 0.01/0.01 0.01/0.01 8.20/50.40 0.01/0.01 0.29/0.30 0.30/0.10

H2S CH3SH DMS DMDS NH3 TMA Acetaldehyde Butyraldehyde

11.33/15.67 1.01/1.25 1.55/1.44 1.41/1.43 489.63/471.43 1.61/0.52 2.11/4.70 0.27/1.18

10.68/11.10 2.79/0.55 1.03/0.72 0.73/0.55 467.63/516.28 3.61/0.34 1.22/3.04 1.19/1.26

28.21/20.69 3.28/3.59 2.59/2.31 2.27/2.45 782.61/662.19 2.98/1.19 4.12/7.29 0.76/2.03

0.45/0.01 0.10/0.20 0.04/0.09 0.07/0.09 5.50/34.92 0.03/0.02 0.37/0.28 0.04/0.08

H2S CH3SH DMS DMDS NH3 TMA Acetaldehyde Butyraldehyde

32.67/40.33 2.01/2.54 2.37/2.09 1.70/1.53 777.37/743.73 3.22/2.12 9.38/14.23 2.31/3.03

55.66/58.02 1.51/2.06 1.37/1.68 2.62/1.23 586.47/522.87 2.80/1.68 5.07/19.86 1.18/2.82

87.32/103.43 4.92/4.12 3.90/4.20 2.98/2.74 1279.15/1169.00 4.31/3.76 15.29/32.87 3.50/4.06

0.08/0.19 0.09/0.07 0.07/0.13 0.06/0.08 10.30/44.84 0.38/0.07 0.76/0.76 0.09/0.30

H2S CH3SH DMS DMDS NH3 TMA Acetaldehyde Butyraldehyde

1.34/1.89 0.19/0.19 0.04/0.06 0.21/0.30 232.98/211.65 0.13/0.04 1.24/1.24 1.64/3.19

1.45/3.44 0.25/0.29 0.34/0.03 0.31/0.95 258.89/320.99 0.30/0.02 1.48/0.32 2.56/3.90

5.42/6.32 0.46/0.67 0.80/0.59 0.58/1.23 319.80/279.50 0.34/0.24 2.11/3.30 3.82/5.60

0.09/0.09 0.01/0.01 0.01/0.01 0.01/0.01 11.80/33.20 0.01/0.01 0.08/0.08 0.02/0.33

H2S CH3SH DMS DMDS NH3 TMA Acetaldehyde Butyraldehyde

5.00/5.53 1.11/1.15 2.26/0.58 1.57/0.61 802.37/420.07 0.34/0.83 2.57/2.97 0.48/0.20

3.68/3.97 0.50/0.75 2.10/0.34 2.59/0.35 793.15/182.18 0.53/0.62 1.79/0.86 2.34/0.12

10.23/11.32 1.74/1.44 4.12/0.92 3.18/1.55 1156.84/1125.50 0.72/1.57 5.60/5.84 0.77/0.45

0.05/0.09 0.02/0.01 0.07/0.01 0.03/0.02 15.20/75.20 0.07/0.07 0.07/0.09 0.02/0.01

H2S CH3SH DMS DMDS NH3 TMA Acetaldehyde Butyraldehyde

1.17/1.04 0.07/0.05 0.08/0.05 0.02/0.02 63.24/63.10 0.02/0.02 0.38/1.03 0.10/0.07

1.89/1.12 0.02/0.03 0.08/0.03 0.12/0.06 24.09/17.02 0.01/0.01 0.15/2.67 0.30/0.05

2.17/1.12 0.10/0.10 0.14/0.16 0.19/0.22 82.10/80.50 0.16/0.23 1.19/3.87 0.17/0.17

0.05/0.05 0.01/0.01 0.01/0.01 0.01/0.02 5.70/4.56 0.01/0.01 0.07/0.01 0.01/0.01

H2S CH3SH DMS DMDS NH3 TMA Acetaldehyde Butyraldehyde

0.23/0.40 0.02/0.03 0.07/0.04 0.03/0.03 62.67/53.33 0.01/0.02 0.18/0.24 0.10/0.14

0.15/0.18 0.01/0.02 0.16/0.02 0.02/0.03 27.16/34.43 0.02/0.02 0.30/0.04 0.15/0.16

0.76/0.60 0.04/0.04 0.19/0.11 0.10/0.20 84.73/90.13 0.12/0.26 0.52/0.49 0.33/0.29

0.05/0.05 0.01/0.01 0.01/0.01 0.01/0.01 5.20/3.50 0.01/0.01 0.02/0.02 0.01/0.03 (continued on next page)

284

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290

Table 5 (continued ) Type

Compound

B-3 site RSCs

H2S CH3SH DMS DMDS NH3 TMA Acetaldehyde Butyraldehyde

NCs CCs a b

Mean

SD

0.17/0.23 0.03/0.06 0.03/0.02 0.03/0.02 60.67/21.97 0.01/0.02 0.19/0.14 0.09/0.10

0.09/0.06 0.02/0.03 0.02/0.01 0.01/0.02 27.97/5.31 0.01/0.01 0.13/0.15 0.02/0.22

Max 0.69/0.42 0.05/0.09 0.03/0.05 0.10/0.20 76.39/30.87 0.06/0.17 0.48/0.33 0.22/0.32

Mina 0.05/0.05 0.01/0.01 0.01/0.01 0.01/0.01 5.50/3.52 0.01/0.01 0.02/0.01 0.01/0.01

Minimum: above detection limit. Summer (JuneeAug.)/Autumn (Oct.eNov.).

5

50

Reduced Sulfur Compounds

45

HS CH SH DMS DMDS

35

4

3

30 25

2

20 15

RSC Concentration(ppb)

H S Concentration(ppb)

40

1

10 5 0

0 A-1

A-2

A-3

A-4

A-5

A-6

B-1

B-2

B-3

Sites 1000

5

Nitrogenous Compounds

NH 4

600

3

400

2

200

1

0

TMA Concentration(ppb)

NH Concentration(ppb)

TMA 800

0 A-1

A-2

A-3

A-4

A-5

A-6

B-1

B-2

B-3

Sites

Acetaldehyde Concentration(ppb)

Carbonyl Compounds

Acetaldehyde Butyraldehyde

14

4 12 10

3

8 2

6 4

1 2 0

Butyraldehyde Concentration(ppb)

5

16

0 A-1

A-2

A-3

A-4

A-5

A-6

B-1

B-2

B-3

Sites Fig. 2. Average concentration of the each of the odorants with respect to the sampling sites.

odorants, but not all of the characteristics of odorants can be evaluated via a comparison of their concentration (Susaya et al., 2011a). It was necessary to utilize the OTV of odorants to exactly trace the cause of the offensive odor because odorants were sensed for different criteria by odor detection levels. Also, we assessed the contribution of odorants more sensitively than the concentration because particularly RSCs are major odorous materials. Hence, we identified the OAV and COI by converting the odor concentration at each research area. We evaluated the OAV, SOAV, and COI using the formula for each odorant (Table 4) considered in this study. Fig. 3 shows the contribution of each odorant at the each sampling site and Fig. 4 shows the differences in COI by functional formula (Table 4). The OAV from RSCs accounted for over 50% (A-1, 65.6%, A2, 53.5%, A-3, 74.7%, A-4, 72.7%, A-6, 66.8%, B-1 72.2%, B-2, 61.7%, B3, 53.7%) at all the sites (except A-5). In the A-5 area, CCs were predominant at 43.5%, RSCs at 36.2%, and NCs at 20.2%. In particularly, the SOAV (172.64) and COI (2.12) at A-4, where food and fish treatment plants were concentrated, was more than 9.5 times that of the other sites, which were in the following order: A-3 (77.50, COI: 1.81), A-6 (44.96, COI: 1.64), A-2 (22.39, COI: 1.22), A-1 (21.30, COI: 1.30), and A-5 (18.18, COI: 1.17). The highest OAV sites were located near the non-point pollution (contaminated river) and food treatment plant, which is known to emit RSCs (Rappert and Muller, 2005). Also, the A-3 site displayed high contributions of RSC, because leather and rubber product plants incorporate an acidification process including a sulfur component. The A-2 and A-5 site near iron production plants exhibited butyraldehyde (31.5% and 39.2%) as the largest odorant, followed by H2S (32.6% and 17.4%). In the residential area, however, the B-1 site indicated extraordinarily high contributions of H2S (58%) from diffusion odorants emitted from the surrounding industrial sites (A-3 and A-4). At the B-2 and B-3 sites, various odorants were similarly distributed and the contribution of butyraldehyde and NH3 detected in the industrial factory (A-2 and A-5) were increased in comparison to the A-1 site. Thus, we should consider the difference in the contributions for each residential area, as the odorous phenomena were particularly influenced by meteorological conditions, especially the northwesterly wind (Susaya et al., 2011a) and individual odorants with dissimilar threshold values could be detected. CH3SH and DMS concentrations were lower than that of H2S (0.01e103.43 ppb) by a factor of 20 and had a range of 0.01e 4.12 ppb and 0.01e4.2 ppb, respectively, but the total contribution of CH3SH and DMS indicated analogous proportions to H2S. Even though the two odorants (CH3SH and DMS) were detected at low concentrations, the results of this frequency analysis suggest the relative dominance of CH3SH and DMS in odor phenomena in the study area. Fig. 4 shows the frequency of COI with respect to each odorant. H2S exhibited the highest COI value of 2.01, followed by CH3SH (1.86), butyraldehyde (1.55), NH3 (1.54), DMS/DMDS (1.41), TMA (1.34), and acetaldehyde (1.25) at the industrial sites. In

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290

285

Fig. 3. The average contribution (%) of odorants at the sampling sites from summer to autumn (except Sep).

contrast, H2S (0.92), DMS (0.52), and butyraldehyde (0.47) exhibited almost undetectable odors at the residential sites and other odors with near zero COI values indicated low contributions as odor constituents (e.g., CH3SH/DMDS (0.38), NH3 (0.37), acetaldehyde (0.25), and TMA (0.15)). The analysis of the COI and OAV patterns of odorants and frequency distribution, asserts that their contribution was important for the inherent odor strength. In particular, the OI at the sites adjacent to non-point pollution sources was more odorous, because the contribution of odorants with a low odor detection threshold may have a cumulative and synergistic effect on malodor phenomena, even at limited occurrence frequencies (O’Neill and Phillips, 1992; Susaya et al., 2011b). Even if individual odorants did not exceed emission criteria, residents living near to the industrial area expressed severe annoyance by identification of unknown odorants or atmospheric phenomena. From a manufacturing viewpoint, the highest odor strength identified in this research area was at the A-4 sites, at which product food and feed productions were present. The offensive odor phenomenon was identified as being influenced by RSCs in the ambient surrounding SSIA and the contribution of odorants were distinguished by classification of industrial activities. However, we cannot entirely assess odor characteristics because odorous nuisances have complex causes including not only the identified odorants, but also other odorants (Volatile Organic Compounds (VOCs) including ketones, alcohols, acids, and acetates). So, we should consider the identification of unknown odorants and the combined effects of factors such as odor dispersion with respect to wind direction pattern and distance to

pollutant emission sources to understand the cause of complex odor problems (Lee et al., 2006; Song et al., 2009). 3.3. Evaluation of the RMR Fig. 5 shows the frequency and mean of OI data at monitoring sites from June to November 2011 excluding September (including rainy days) and we assessed OI (2) with annoyance criteria (Nimmermark et al., 2005). The reliability analysis yielded a Cronbach’s alpha index from 0.564 to 0.636 in the SSIA and 0.651 to 0.717 in the residential area, so almost all sites showed satisfactory reliability (0.60), with the residential area more reliable than the odor emission area. The OI values in the industrial area exhibited the highest levels at A-4 (2.04, n ¼ 32760), and likewise with COI (2.12), followed by A-3 (1.80, COI: 1.81, n ¼ 31920), A-6 (1.71, COI: 1.64, n ¼ 30240), A-2 (1.53, COI: 1.22, n ¼ 31035), A-5 (1.52, COI: 1.17, n ¼ 31920), and A-1 (1.38, COI: 1.30, n ¼ 31006) whereas, in the residential area, the OI values occurred in the order B-3 (0.63, COI: 0.24, n ¼ 30240), B-1 (0.60, COI: 0.54, n ¼ 31080), and B-2 (0.50, COI: 0.38, n ¼ 32755). Food and beverage production facilities are located in the A-4 and found to be the most malodorous sources. In general, the OI values also observed by residents were higher values than the COI values measured by instrumental analysis and the difference in OI was not distinguishable from the classification of industrial activities comparison using COI. Especially, the A-2 and A-5 sites among the industrial sites showed strong differences in OI values, and B-3 at the residential area had a higher value, 2.8 times greater than COI. Table 6 shows the frequency characteristics of the odor description and complaint

286

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290 90

90

H2S

70

Frequency

60 50 40

CH3SH

80

A-1 A-2 A-3 A-4 A-5 A-6 B-1 B-2 B-3

A-1 A-2 A-3 A-4 A-5 A-6 B-1 B-2 B-3

70 60

Frequency

80

50 40

30

30

20

20

10

10

0

0

0~1

1~1.5

1.5~2

2~2.5

2.5~3

3~3.5

3.5~4

4~5

0~1

1~1.5

Computed odor intensity (COI)

DMS A-1 A-2 A-3 A-4 A-5 A-6 B-1 B-2 B-3

70 60

Frequency

2.5~3

3~3.5

3.5~4

50 40

80

4~5

DMDS

70

A-1 A-2 A-3 A-4 A-5 A-6 B-1 B-2 B-3

60

Frequency

80

50 40

30

30

20

20

10

10 0

0 0~1

1~1.5

1.5~2

2~2.5

2.5~3

3~3.5

3.5~4

0~1

4~5

1~1.5

1.5~2

2~2.5

2.5~3

3~3.5

3.5~4

4~5

Computed odor intensity (COI)

Computed odor intensity (COI) 90

90

NH3 A-1 A-2 A-3 A-4 A-5 A-6 B-1 B-2 B-3

70 60 50 40

TMA

80

A-1 A-2 A-3 A-4 A-5 A-6 B-1 B-2 B-3

70 60

Frequency

80

Frequency

2~2.5

90

90

50 40

30

30

20

20

10

10

0

0

0~1

1~1.5

1.5~2

2~2.5

2.5~3

3~3.5

3.5~4

4~5

0~1

1~1.5

Computed odor intensity (COI)

1.5~2

2~2.5

2.5~3

3~3.5

3.5~4

4~5

Computed odor intensity (COI)

90

90

Acetaldehyde A-1 A-2 A-3 A-4 A-5 A-6 B-1 B-2 B-3

70 60 50 40

A-1 A-2 A-3 A-4 A-5 A-6 B-1 B-2 B-3

70 60 50 40

30

30

20

20

10

10

0

Butyraldehyde

80

Frequency

80

Frequency

1.5~2

Computed odor intensity (COI)

0 0~1

1~1.5

1.5~2

2~2.5

2.5~3

3~3.5

3.5~4

Computed odor intensity (COI)

4~5

0~1

1~1.5

1.5~2

2~2.5

2.5~3

3~3.5

3.5~4

4~5

Computed odor intensity (COI)

Fig. 4. Frequency distribution of all sampling sites with respect to the computed odor intensity (COI) by instrument analysis from Jun to Nov 2011.

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290

pollution was a serious problem and experienced irritations due various malodor sources. Hence, an odor emission standard may be required in terms of human health as a reinforcement of the criteria and regulations of maximum concentration, and we need to compare guidelines with various international criteria. There is general agreement on which odors are experienced as unpleasant, e.g., odors that are pungent (NH3, acetaldehyde), rotten eggs (H2S), decaying cabbage (CH3SH and DMS), stinking (garbage wastes), fishy (TMA), and rancid (Butyric acid) (Yuwono and Lammers, 2004). The odor descriptions classified five odorous characteristics (Sour, Enamel, Musty, Disgusting, Sewage, and Unknown). The frequency of the ‘Rotten’ smell was higher at 25.3% than any other smell, followed by ‘Pungent’ (20.9%), ‘Other’ (16.8%), ‘Fishy’ (15.4%), ‘Burned’ (12.8%), and ‘Fruity’ (8.7%). Here, we note that unidentified malodor (‘Other’) may have a cumulative influence on odor irritation or annoyance and the sensation of panelists was susceptible to mixed smells because odorants emitted by various emission sources could diffuse to give a complex smell in ambient air. Fig. 6 shows the frequency and mean of OI with respect to odor description. The result of frequency by OI (2) for ‘Fruity’, ‘Fishy’, ‘Rotten’, ‘Pungent’, ‘Burned’, and ‘Other’ were 17%, 24.2%, 28.7%, 17.1%, 12.9%, and 35.9% in the industrial area, respectively, while in the residential area, they were 12%, 10.4%, 9.1%, 11%, 15.8%, and 12.7%, respectively. The ‘Rotten’ results in the industrial also exhibited a strong compatibility with OI, but the frequency and OI in the residential area declined. Conversely, the OI proportion of ‘Other’ and ‘Burned’ increased in the residential area. In other words, inhabitants who participated in odor monitoring may not actually detect certain malodorous characteristics and they noticed more simple ‘Burned’ components rather than other irritant sensations. Such a discrepancy, in which the ‘Other’ frequency was highest, may be construed as a lower odor concentration in comparison to the offensive odor source area, and they could be limited to the evaluation of major odor degrees accurately over a small area. On the other hand, this different outcome shows the drawbacks of direct olfaction using questionnaires owing to the lack of impartiality of some residents, as well as low motivation (Nicolas et al., 2010). Therefore, we should verify the validity of the correlation analysis between the instrumental analysis data and RMR. From the results from the RMR, the OI values were higher than the COI values and even the overall degree of annoyance exceeded 50%. The ‘Rotten’ odor, which was known to be caused by RSCs,

840

720

Frequency

600

480

360 0 : None odor 1 : Slight 2 : Noticeable 3 : Strong 4 : Very strong 5 : Unbearable Annoyance criteria

240

120

0 A-1

A-2

A-3

A-4

A-5

A-6

B-1

B-2

287

B-3

Sites Fig. 5. Frequency distribution of the odor intensity (OI) by RMR method with respect to the sampling sites from Jun to Nov 2011.

ratio by residents with respect to the sites. The annoyance level (OI (2)) at A-4 was highest at 66.5%, and in the A-1 area, it was lowest at 41.4%. The overall degree of annoyance was 51.9% in the industrial area, so we could interpret the occurrence of a serious odor phenomenon sensed by the residents. The difference arises owing to participants tending to depend on a subjective sense being influenced by the odor type and intensity and the odor strength being irregular due to local winds. Hence, we could confirm that RMR was significant in terms of inherent odor strength by comparison with the measured concentration or COI. Furthermore, it is reasonable to suspect that phenomena such as odor malodor are not simple enough to be diagnosed by OI and OAV because the actual OI in complex industrial areas could be a combined OI due to the presence of other odorants with different odor parameters (Blanes-Vidal et al., 2012; Susaya et al., 2011a). It is worthwhile discussing these discrepancies between environmental criteria and direct olfaction, and therefore we need to consider legislative criteria for odorant control. Nevertheless, the overall measured concentrations during this study period remained within air quality guidelines, other than a few of the maximum concentrations. However, the residents living nearby still perceived that odorous

Table 6 An overall summary of odor characteristics from direct olfactory method observed by residents from June to November 2011. Monitoring sites

N

Odor description

Odor intensity (OI)

Frequency (Ratio, %)

Mean  SD

Complaint index (OI < 2)

Industrial areas A-1 31006 A-2 31035 A-3 31920 A-4 32760 A-5 31920 A-6 30240 Average (%) Resident areas B-1 31080 B-2 32755 B-3 30240 Average (%)

Burned

Other (including unknown)

(OI  2)

Fruity

Fishy and disgusting

Rotten-decayed

Pungent

Ratio (%)

1221 2849 3382 5655 1748 1764 2762

(3.9) (9.0) (10.6) (17.3) (5.5) (5.8) (8.7)

5365 2072 6726 9087 2584 3528 4894

(17.3) (6.7) (21.1) (27.7) (8.1) (11.7) (15.4)

6737 6919 9728 10023 6118 8316 7973

(21.7) (22.3) (30.5) (30.6) (19.2) (27.5) (25.3)

8991 7363 5054 2613 7600 7668 6548

(29.0) (23.7) (15.8) (8.0) (23.8) (25.4) (20.9)

4625 4144 3344 1716 8056 2304 4032

(14.9) (13.4) (10.5) (5.2) (25.2) (7.6) (12.8)

4070 7733 3686 3666 5814 6660 5272

(13.1) (24.9) (11.5) (11.2) (18.2) (22.0) (16.8)

1.38 1.53 1.80 2.04 1.52 1.71 1.67

      

3.88 2.30 2.24 4.89 2.45 1.66 1.99

58.9 54.9 41.8 33.5 56.8 42.9 48.1

41.1 45.1 58.2 66.5 43.2 57.1 51.9

1221 2184 828 1411

(3.9) (6.7) (2.7) (4.4)

2516 2842 3168 2844

(8.1) (8.7) (10.5) (9.1)

8214 7059 3780 6351

(26.4) (21.5) (12.5) (20.2)

4514 4212 2484 3737

(14.5) (12.9) (8.2) (11.9)

6808 7371 10980 8386

(21.9) (22.5) (36.3) (26.9)

7807 9087 9000 8631

(25.1) (27.7) (29.8) (27.5)

0.60 0.50 0.63 0.68

   

0.48 0.55 0.62 0.53

80.4 86.2 81.5 82.7

19.6 13.8 18.5 17.3

288

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290

log(SOC)

6

(A)

5

4

3

2

1

-0.6

-0.4

-0.2

y = 1.9834x+1.8589 r = 0.465 0.0

0.2

0.4

0.6

0.8

1.0

log(Odor Intensity)

log(SOAV)

5

(B)

4

3

2 Fig. 6. Comparison of frequency and odor intensity (OI) with respect to the odor description at all monitoring sites from Jun to Nov 2011.

1

y = 2.4741x+1.4442 r =0.720 -0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

log(Odor Intensity)

7

SCOI

tended to have a similar frequency to the distribution of the odor concentration in the industrial area. In contrast, the cause of offensive malodors in the residential area was indicated ‘Burned’ and ‘Other’ smells in residential area. We recognize that individual sensitivity to the OI can vary significantly when odorants are emitted into the ambient air, and individuals may or may not perceive an odor (Yuwono and Lammers, 2004). This difference could suggest that other odorous compounds, which were not measured, may be continuously emitted in the SSIA. In addition, people tend to perceive what they regard as unacceptable odor types, and a major offensive odor emitted from an industrial area can become an odor problem. Consequently, we need to carefully assess odorous nuisances resulting from various odor sources, and we must invest in various types of facilities at the sampling sites and measure a variety of odorants emitted from the industrial complex. Moreover, we should evaluate odor annoyance using indirect (instrumental) methods and direct methods based on steady social participation over a long period to resolve odor problems and health risks for people who live near the industrial complex.

(C)

6

5

4

3

2

y = 3.355x+2.2178 r = 0.629

1

3.4. Correlation analysis of odor intensity These converted index values were calculated on a logarithmic scale to compare with the OI values evaluated by RMR and the correlation analysis results are shown in Fig. 7. Fig. 7A shows a graph that compares the correlation result between OI by RMR and SOC in terms of logarithmic scales. The Pearson’s correlation coefficient (r) from the analysis indicated values of 0.465 (P ¼ 6.1E-

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

log(Odor Intensity) Fig. 7. The result of the correlation analysis between odor intensity (OI) from RMR and computed indices (SOC, SOAV, and SCOI) from instrumental analysis.

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290

04), but this was lower than other reported values of 0.671 (P ¼ 1.9E-10) (Kim and Park, 2008). They analyzed the relationship between direct measurements (olfactometry) using an air dilution sensory test that quantifies the odor index values in terms of dilution to threshold (D/T) ratios derived by the air dilution sensory (ADS) (Kim, 2011) and indirect (instrumental) detection. The static dilution of odor samples for the ADS test was made in a stepwise manner by mixing original samples with odorless air using a 3 L odor bag made of polyethylene telephtalate film. D/T ratios of pure air volume required to dilute sampled odorous air to an odor-free threshold point by the stipulated method (Kim and Park, 2008; Mao et al., 2006). In particular, this analysis method is acknowledged as a meaningful approach in the assessment of odor strength. So, through this result, only the analysis of the SOC correlation values lack the ability to objectively identify odor problems. The results expressed in terms of SOAV on a logarithmic scale are shown in Fig. 7B. The Pearson correlation coefficient (r) indicated a relatively high value of 0.720 (P ¼ 7.72E-12) compared to the SOC coefficient value. Kim and Park (2008) also concluded that the SOQ term exhibited an exceptionally strong correlation to the D/T ratio (r ¼ 0.866, P ¼ 2.74E-22). We assume that the high correlation value is due to the SOAV index being calculated using OTV so that the sensing level of odorous component can be assessed. Likewise, we evaluated the correlation of the OI values from the RMR with the SCOI values from the instrumental analysis. The Pearson correlation coefficient (r) also indicated moderate correlation with a value of 0.629 (P ¼ 6.94E-09), although this is lower compared to the SOAV correlation values. The overall results of the correlation coefficients at the SSIA showed a moderately high correlation distribution range (from 0.465 to 0.720, n ¼ 50). Therefore, we need to consider continuous monitoring activities for residents to identify the cause of odorous phenomena and to establish management countermeasures in industrial complexes. 4. Conclusions We determined the concentrations of the odorants by instrumental analysis and resident monitoring records of direct olfactory observations. The highest odor strength was identified at the A-4 site at which product food and feed production facilities were located. The concentrations of RSCs (especially H2S) were significantly higher than the other odorous compounds, and the contributions of odorants were different depending on the facilities located in each site. Especially, the RSCs contribution was very high in some residential sites while the other residential sites showed different results. The OI values obtained from RMR were higher than the COI values obtained from instrumental analysis. This suggests that a clear guideline is necessary to determine the emissions standards in Korea. A strong correlation was found between the OI by RMR and COI/SOAV by instrumental analysis. The RMR method was effective to understand the characteristics of odorants emissions from the SSIA. From this study, an odor problem was found to be serious in the SSIA; therefore, further research is necessary to find major odor sources and to control odor emissions. Acknowledgment This work was supported by the second stage of the Brain Korea 21 Project in 2012 and by Pusan National University Research Grant. Nomenclature ADS CCs CI

air dilution sensory carbonyl compounds confidence interval

COI DMS DMDS D/T GC HPLC MDL NCs OAV OI ODS OTV PFPD QA RMR RSCs RSE SCOI SOAV SOC SOQ SSIA TMA VOCs

289

computed odor intensity dimethyl sulfide dimethyl disulfide dilution-to-threshold ratio gas chromatograph high performance liquid chromatography minimum detection limit nitrogenous compounds odor activity value odor intensity octadecyl silane odor threshold value pulsed flame photometric detector quality assurance resident monitoring records reduced sulfur compounds relative standard error sum of computed odor intensity sum of odor activity value sum of odor concentration sum of odor quotient Sa-Sang industrial area trimethylamine volatile organic compounds

References Aatamila, M., Verkasalo, P.-K., Korhonen, M.J., Viluksela, M.-K., Pasanen, K., Tiittanen, P., Nevalainen, A., 2012. Odor annoyance near waste treatment centers: a population-based study in Finland. Journal of the Air & Waste Management Association 60, 412e418. Blanes-Vidal, V., Suh, H., Nadimi, E.S., Løfstrøm, P., Ellermann, T., Andersen, H.V., Schwartz, J., 2012. Residential exposure to outdoor air pollution from livestock operations and perceived annoyance among citizens. Environment International 40, 44e50. Bundy, D.S., 1992. Odor issues with wastes. In: National Livestock Poultry and Aquaculture Waste Management. American Society of Agricultural Engineers (ASAE) Publication, St. Joseph, Michigan, pp. 288e292. Capelli, L., Sironi, S., Del Rosso, R., Centola, P., Rossi, A., Austeri, C., 2011. Olfactometric approach for the evaluation of citizens’ exposure to industrial emissions in the city of Terni, Italy. Science of the Total Environment 409, 595e603. Chen, Y.C., Bundy, D.S., Hoff, S.J., 1999. Development of the relationship between odor intensity and threshold dilution ratio for swine units. Journal of the Air & Waste Management Association 49, 1082e1088. Chien, Y.C., Uang, S.N., Kuo, C.T., Shih, T.S., Jen, J.F., 2000. Analytical method for monitoring trimethylamine using solid phase micro-extraction and gas chromatography-flame ionization detection. Analytica Chimica Acta 419, 73e79. Choi, Y.J., Jeon, E.C., Kim, K.H., 2005. The emission characteristics of reduced sulfur compounds in the ban wall industrial complex. Journal of Korean Society for Atmospheric Environment 21, 515e524. Emerson, C.W., Rajagopal, R., 2004. Measuring toxic emissions from landfills using sequential screening. Computers Environment and Urban Systems 28, 265e 284. Feilberg, A., Liu, D., Adamsen, A.P.S., Hansen, M.J., Jonassen, K.E.N., 2010. Odorant emissions from intensive pig production measured by online proton-transferreaction mass spectrometry. Environmental Science & Technology 44, 5894e 5900. Gallego, E., Soriano, C., Roca, F.X., Perales, J.F., Alarcon, M., Guardino, X., 2008. Identification of the origin of odour episodes through social participation, chemical control and numerical modeling. Atmospheric Environment 42, 8150e8160. Gostelow, P., Parsons, S.A., Stuetz, R.M., 2001. Odour measurements for sewage treatment works. Water Research 35, 579e597. Heaney, C.D., Wing, S., Campbell, R.L., Caldwell, D., Hopkins, B., Richardson, D., Yeatts, K., 2011. Relation between malodor, ambient hydrogen sulfide, and health in a community bordering a landfill. Environmental Research 111, 847e852. Herr, C., Zur Nieden, A., Bödeker, R.H., Gieler, U., Eikmann, T.F., 2003. Ranking and frequency of somatic symptoms in residents near composting sites with odor annoyance. International Journal of Hygiene and Environmental Health 206, 61e64. Jiang, J., Coffey, P., Toohey, B., 2006. Improvement of odor intensity measurement using dynamic olfactometry. Journal of the Air & Waste Management Association 56, 675e683. Kabir, E., Kim, K.-H., 2010. An on-line analysis of 7 odorous volatile organic compounds in the ambient air surrounding a large industrial complex. Atmospheric Environment 44, 3492e3502.

290

H.-D. Lee et al. / Atmospheric Environment 74 (2013) 277e290

Kim, K.-H., 2011. The averaging effect of odorant mixing as determined by air dilution sensory tests: a case study on reduced sulfur compounds. Sensors 11, 1405e1417. Kim, K.-H., Park, S.-Y., 2008. A comparative analysis of malodor samples between direct (olfactometry) and indirect (instrumental) method. Atmospheric Environment 42, 5061e5070. Kim, K.-H., Choi, Y.J., Jeon, E.C., Sunwoo, Y., 2005. Characterization of malodorous sulfur compounds in landfill gas. Atmospheric Environment 39, 1103e1112. Kim, K.-H., Jeon, E.-C., Choi, Y.-J., Koo, Y.-S., 2006. The emission characteristics and the related malodor intensities of gaseous reduced sulfur compounds (RSC) in a large industrial complex. Atmospheric Environment 40, 4478e4490. Kim, K.-H., Pal, R., Ahn, J.-W., Kim, Y.-H., 2009. Food decay and offensive odorants: a comparative analysis among three types of food. Waste Management 29, 1265e 1273. KMOE, 2007a. Offensive Odorant Prevention Law. Korean Ministry of Environment (KMOE) (in Korean). KMOE, 2007b. Odorant Management Handbook. Korean Ministry of Environment (KMOE) (in Korean). Lee, S., Xu, Q., Booth, M., Townsend, T.G., Chadik, P., Bitton, G., 2006. Reduced sulfur compounds in gas from construction and demolition debris landfills. Waste Management 26, 526e533. Lomans, B.P., Pol, A., Camp, H.J.M.O.d., 2002. Microbial cycling of volatile organic sulfur compounds in anoxic environments. Water Science and Technology 45, 55e60. Mackie, R.I., Stroot, P.G., Varel, V.H., 1998. Biochemical identification and biological origin of key odour components in livestock waste. Journal of Animal Science 7, 1331e1342. Mao, I.-F., Tsai, C.-J., Shen, S.-H., Lin, T.-F., Chen, W.-K., Chen, M.-L., 2006. Critical components of odors in evaluating the performance of food waste composting plants. Science of the Total Environment 370, 323e329. Micone, P.G., Guy, C., 2007. Odor quantification by a sensor array: an application to landfill gas odors from two different municipal waste treatment works. Sensors and Actuators B-Chemical 120, 628e637. Mushiroda, T., Yokoi, T., Takahara, E., Nagata, O., Kato, H., Kamataki, T., 1999. Sensitive assay of trimethylamine N-oxide in liver microsomes by headspace gas chromatography with flame thermionic detection. Journal of Chromatography B 734, 319e323. Nagata, Y., 2003. Odor Intensity and Odor Threshold Value. Bulletin of Japan, Environmental Sanitation Center, Japan, pp. 17e25. Nicolas, J., Cors, M., Romain, A.-C., Delva, J., 2010. Identification of odour sources in an industrial park from resident diaries statistics. Atmospheric Environment 44, 1623e1631. Nimmermark, S.A., Jacobson, L.D., Schmidt, D.R., Gay, S.W., 2005. Predictions by the odor from feedlots, setback estimation tool (OFFSET) compared with observations by neighborhood monitors. Journal of the Air & Waste Management Association 55, 1306e1314. O’Neill, D.H., Phillips, V.R., 1992. A review of the control of odour nuisance from livestock buildings: part 3, properties of the odorous substances which have

been identified in livestock wastes or in the air around them. Journal of Agricultural Engineering Research 53, 23e50. Parker, D.B., Gilley, J., Woodbury, B., Kim, K.-H., Galvin, G., Bartelt-Hunt, S.L., Li, X., Snow, D.D., 2013. Odorous VOC emission following land application of swine manure slurry. Atmospheric Environment 66, 91e100. Rappert, S., Muller, R., 2005. Odor compounds in waste gas emissions from agricultural operations and food industries. Waste Management 25, 887e907. Rosenfeld, P.E., Henry, C.L., Benett, D., 2001. Wastewater dewatering polymer affect on biosolids odour emissions and microbial activity. Water Environment Research 73, 363e367. Seo, S.-G., Ma, Z.-K., Jeon, J.-M., Jung, S.-C., Lee, W.-B., 2011. Measurements of key offensive odorants in a fishery industrial complex in Korea. Atmospheric Environment 45, 2929e2936. Smet, E., Lens, P., Van Langenhove, H.V., 1998. Treatment of waste gases contaminated with odorous sulfur compounds. Critical Reviews in Environmental Science and Technology 28, 89e117. Song, S.K., Shon, Z.H., Kim, K.H., 2009. Photochemical oxidation and dispersion of gaseous sulfur compounds from natural and anthropogenic sources around a coastal location. Atmospheric Environment 43, 3015e3023. Sucker, K., Both, R., Bischoff, M., 2008. Odor frequency and odor annoyance. Part 1: assessment of frequency, intensity and hedonic tone of environmental odors in the field. International Archives of Occupational Environmental Health, 671e 682. Susaya, J., Kim, K.-H., Phan, N.-T., Kim, J.-C., 2011a. Assessment of reduced sulfur compounds in ambient air as malodor components in an urban area. Atmospheric Environment 45, 3381e3390. Susaya, J., Kim, K.-H., Chang, Y.-S., 2011b. Characterization of major offensive odorants released from lake sediment. Atmospheric Environment 45, 1236e1241. Toda, T., Obata, T., Obolkin, V.A., Potemkin, V.L., Hirota, K., Takeuchi, M., Arita, S., Khodzher, T.V., Grachev, M.A., 2010. Atmospheric methanethiol emitted from a pulp and paper plant on the shore of Lake Baikal. Atmospheric Environment 44, 2427e2433. Trabue, S.L., Anhalt, J.C., Zahn, J.A., 2006. Bias of tedlar bags in the measurement of agricultural odorants. Journal of Environmental Quality 35, 1668e1677. Trabue, S., Scoggin, K., McConnell, L., Maghirang, R., Razote, E., Hatfield, J., 2011. Identifying and tracking key odorants from cattle feedlots. Atmospheric Environment 45, 4243e4251. Tsao, Y.-C., Wu, C.-F., Chang, P.-E., Chen, S.-Y., Hwang, Y.-H., 2011. Efficacy of using multiple open-path Fourier transform infrared (OP-FTIR) spectrometers in an odor emission episode investigation at a semiconductor manufacturing plant. Science of the Total Environment 409, 3158e3165. Yu, Y.J., Chea, H.Y., Yu, J.H., Huh, E.J., Kim, M.G., 2009. The emission characteristics of aldehydes from industrial complex in Busan. Korean Journal of Odor Research and Engineering 8, 188e195. Yuwono, A.S., Lammers, P.S., 2004. Odor pollution in the environment and detection instrumentation. Agricultural Engineering International: The CIGR Journal of Scientific Research and Development, Invited Overview Paper 6, 1e33. Zhang, Z., Xu, K., Xing, Z., Zhang, X., 2005. A nanosized Y2O3-based catalytic chemiluminescent sensor for trimethylamine. Talanta 65, 913e917.