Abstract
Indoor air quality has become a more prominent concern since the arrival of the COVID-19 pandemic. Manufacturing industries have always been prone to occupational health risks, which depend on the dynamics of the production shop floors. The furniture industry is one of these sectors with a unique work environment. Although a typical furniture manufacturing facility involves physical, chemical, and noise pollution-producing elements, this industry has been studied relatively less for indoor air quality and noise-related risks. This study investigated nine furniture manufacturing organizations’ indoor air quality and noise pollution levels through comprehensive quantitative techniques. The results of the measurements were compared against reference values set by specific guidelines to explore the degree of occupational health risk associated with the World Health Organization’s (WHO) suggested levels. Repetitive measurements from five pre-designated workstations were taken at each facility. The study’s results indicated that organization size and department were significant factors for PM 2.5 and HCHO parameters, while only department type was substantial for noise exposure levels. However, across all departments and organization sizes, LAeq noise levels were below the safety threshold of 85 dB(A). Most organizations presented a lack of proper use of personal protective equipment and poor ventilation across shop floors.
Download PDF
Full Article
Investigation of Indoor Noise Pollution Level and Air Quality of Furniture Manufacturers
Cagatay Tasdemir,a,* Yalcin Yildirim,b Mesut Uysal,a Naile Angin,a and Murat Ertas a
Indoor air quality has become a more prominent concern since the arrival of the COVID-19 pandemic. Manufacturing industries have always been prone to occupational health risks, which depend on the dynamics of the production shop floors. The furniture industry is one of these sectors with a unique work environment. Although a typical furniture manufacturing facility involves physical, chemical, and noise pollution-producing elements, this industry has been studied relatively less for indoor air quality and noise-related risks. This study investigated nine furniture manufacturing organizations’ indoor air quality and noise pollution levels through comprehensive quantitative techniques. The results of the measurements were compared against reference values set by specific guidelines to explore the degree of occupational health risk associated with the World Health Organization’s (WHO) suggested levels. Repetitive measurements from five pre-designated workstations were taken at each facility. The study’s results indicated that organization size and department were significant factors for PM 2.5 and HCHO parameters, while only department type was substantial for noise exposure levels. However, across all departments and organization sizes, LAeq noise levels were below the safety threshold of 85 dB(A). Most organizations presented a lack of proper use of personal protective equipment and poor ventilation across shop floors.
DOI: 10.15376/biores.19.2.3571-3596
Keywords: Furniture industry; Indoor air quality; Indoor pollutants; Noise measurement; Occupational health; Ergonomics
Contact information: a: Department of Forest Industry Engineering, Bursa Technical University, Bursa, Türkiye; b: Department of Landscape Architecture, Bursa Technical University, Bursa, Türkiye;
* Corresponding author: cagatay.tasdemir@btu.edu.tr
INTRODUCTION
The furniture industry is an essential sector with a worldwide economic contribution (Abu et al. 2019). The USA, China, Canada, and Italy lead the furniture industry globally (Pirc and Vlosky 2010). In addition, many local manufacturers in other Asian and European countries serve the furniture market. Türkiye is an essential player in the global furniture market, with an export value of more than 554 million USD annually (Çınar 2005; IMOS 2019). The country is a decisive force in regional industrial dynamics due to its forest product stocks and geo-strategic location between the East and West. Türkiye is an essential supplier of some tree species, such as pine, beech, and fir, which are used for furniture manufacturing (Coşkun 2019).
One of the most significant value-generating furniture manufacturer clusters in Türkiye is in the Bursa-Inegol region. The city is located on the historical Silk Road and has been famous for its furniture among traders using this route. In the last 15 to 20 years, in the Inegol Region, the number of small-, medium-, and large-scale enterprises has increased; today, there are around 587 furniture exporting companies (Araz and Yaşar 2020). Within Türkiye’s furniture industry, beech, hornbeam, poplar, and pine are the most commonly utilized wood species for upholstered furniture manufacturing, while walnut, oak, and some other hardwood species are preferred for furniture items made of solid wood.
Furniture has always been a reflection of culture and lifestyle for societies. The furniture industry has been directly affected by many global and regional social and natural factors such as wars, population trends, natural disasters, technological improvements, pandemics, and, last but not least, resource scarcity. For instance, after the Second World War, there was a worldwide shortage in the supply of raw wood and wood-based materials, and the furniture industry had a difficult time, leading to the emergence of alternative furniture designs (Pirc and Vlosky 2010). Similarly, the COVID-19 pandemic started to shake the world in 2019 and directly affected the furniture industry’s manufacturers and consumers (Ratnasingam et al. 2020). Due to the pandemic, people have been forced to adopt the home-office working model and started to spend more time at home. Therefore, the demand for furniture has increased out of the ordinary (Pirc Barčić et al. 2021). The increasing demand has caused furniture workers to work indoors over capacity for more extended hours, which has meant longer exposure to working environment conditions. The furniture industry produces semi-finished or finished products from wood by cutting, sanding, mowing, chipping, fibering, bonding, pressing, steaming, drying, and impregnation processes (Smardzewski 2015). During these processes, physical injuries and other occupational hygiene and health-threatening risks may occur due to exposure to dust, chemical gas, noise, vibration, and thermal discomfort. Additionally, indoor air quality risks have recently been more prominent and attention-grabbing due to the COVID-19 pandemic (Abouleish 2021; Tian et al. 2021).
Indoor air quality should include key aspects to provide a healthy and comfortable working environment (Persily 2015). These features are stated in Standard 62 of the American Society of Heating, Refrigerating, and Air-Conditioning (ASHRAE). Indoor air pollutants could be classified as gases and particulates (Batterman and Peng 1995). The well-known gas pollutants are CO, CO2, formaldehyde, volatile organic components (VOC), O3, NO2, and SO2. Particle matter (PM) is a tiny solid pollutant, usually originating from the external environment or dust generated during manufacturing processes (Alptekin and Çelebi 2015). Depending on emission sources and air conditions, particle matter’s density and chemical compositions change. Because fine (PM 2.5) and coarse (PM 10) particles come from different sources and show different physical and chemical properties, PM 2.5/PM 10 ratios can provide important information about the source, formation, and effects of particles on human health (Bozkurt 2018). Past research revealed that dust exposure during wood processing mainly occurs at the sanding station, and it was reported that furniture dust negatively affects lung functions and causes serious diseases such as asthma (Mikkelsen et al. 2002; Jacobsen et al. 2008). A survey study conducted among 30 selected furniture factories in the Southeast Asian region showed that total inhalable dust particles were less than 10 µm in diameter, and their concentration was less than 25% by weight (Ratnasingam et al. 2010).
In addition to gases and particulates, noise is another occupational risk category that must be addressed. A few studies examined the effects of manufacturers and their facilities on occupational health, including safety, respiratory, and noise at various scales. Lie et al. (2016) reported a literature survey covering almost 700 articles and delved into roughly 200 of them to understand whether noise exposure results in a hearing decrease or loss among workers in terms of occupational health context (Lie et al. 2016). The study found that men tend to experience more hearing loss. On top of that, besides noise, some other factors, such as vibration and chemical substances, played essential roles in this concern. Bharwana et al. (2019) conducted a study with fifty workers of iron furniture manufacturers to understand occupational risks (Bharwana et al. 2019). The authors utilized a survey and reported high dust, heat, and noise exposure as notable hazards in small-scale manufacturers. Ntalos and Papadopoulos (2005) observed a similar trend of noise exposure in furniture manufacturing firms. The study found that all noise measurements were above 85 dB(A), which is considered unhealthy working conditions, particularly when exposed for more than 8 hours.
Considering more noise-specific studies, Malkin et al. (2005) performed a study in seven wood pallet production companies to assess noise exposure. The authors measured the noise levels and found that noise levels associated with machines and machine-related activities in each site were above 90 dB(A). Filipe et al. (2014) conducted a study in fourteen furniture factories to examine noise exposure in Brazil. The noise levels ranged around 50 dB(A) during 8-hr measurements. The study reported that measurements were above accepted noise levels of Brazilian regulations. Guarnaccia et al. (2013) performed an experimental study in wood production firms to understand single-source noise level exposure when the workers were in their working routine. The study found that various specific wood processing equipment, including band saws, circular saws, and nail guns, are responsible for high noise levels and concluded that mitigation policies should be taken for frequently utilized higher noise-level equipment.
Durcan and Burdurlu (2018) studied wood materials from a more specific perspective by evaluating the MDF made of Lombardy Poplar at various thicknesses ranging from 6 to 30 mm within a twenty-minute production time (Durcan and Burdurlu 2018). The study results showed that noise levels increased by up to 9 dB(A) with increasing levels of thickness. In a more recent study, Fidan et al. (2020) performed a study that examined noise levels only in lumber processing sections of forest product manufacturers (Fidan et al. 2020). The study was conducted in 17 work areas with a 5-second sequence of three-minute sampling. The measurements were analyzed and interpreted as some equipment, including a vertical wood band sawmill, were operating at higher noise levels than other machinery. The study suggested the adoption of protective precautions depending on the requirements of each case within manufacturing facilities. However, these past studies primarily focused on noise measurements of specific machinery and equipment in offices, stores, and shop floors and did not address the co-existence of other occupational health threats from a holistic perspective. Most of these studies also did not evaluate the situation comparatively for furniture companies of various scales. Furthermore, some past studies solely focused on shop-floor activities and ignored administrative offices and warehouses. In addition to comprehensive indoor air quality measurements, this study investigated ambiance noise levels through quantitative measurement techniques and compared them with reference values set by the World Health Organization to take a snapshot of the relatively less charted territory, furniture manufacturing facilities located in Inegol-Bursa.
The motivations behind this study were multifaceted, stemming from the increasing concern for occupational health within the manufacturing sector, particularly in settings prone to air pollutants and noise exposure such as the furniture manufacturing industry. This industry’s unique intersection of chemical usage, wood dust generation, and machinery operation presents significant health risks, meriting a detailed investigation. A notable gap in the literature was identified: the lack of comprehensive research evaluating indoor air quality and noise pollution across different organizational scales within the furniture manufacturing sector. This gap was particularly pressing in the wake of the COVID-19 pandemic, which has heightened awareness around the importance of indoor air quality for public and occupational health and altered industrial work dynamics, potentially intensifying exposure to indoor pollutants. Furthermore, there existed a pressing need for data-driven recommendations to inform the development of targeted interventions and policies aimed at mitigating health risks in this industry.
Therefore, the objectives of this study were to 1) determine and compare indoor air quality and ambient noise levels of the small, medium, and large-scale organizations within the furniture industry and 2) identify and discuss the chronic indoor air pollution and noise-associated risks along with underlying factors of these risks from a holistic perspective in terms of occupational health and safety.
EXPERIMENTAL
The study methodology employed could be summarized in three phases: 1) the identification and selection of participating firms, 2) the determination of measurement parameters and measurement locations, and 3) data collection, processing, and reporting.
The selection of the participating firms, which were subjected to ambient air quality and noise level measurements, was completed systematically. A target population of 350 firms involving small-, medium-, and large-sized enterprises that meet the corresponding size classification criteria of the Ministry of Industry and Technology of the Republic of Türkiye were identified by using the membership database of the Association of Inegol Furniture Manufacturers (IMOS), a regional authority awarded with ECEI Bronze Label by European Secretariat in 2014. Micro-sized enterprises were excluded from the population, and the remaining population size was clustered into three size categories: small-, medium-, and large-sized enterprises. Then, three firms from each category were selected through a judgmental sampling procedure (Duignan 2016). A total of nine companies from all size categories were identified as main participants of the study, and nine more firms (three from each size category) were also selected and contacted as backup facilities or data sources, as illustrated in Fig. 1. Confidentiality and cooperation agreements were signed with all of the participating firms. Therefore, company names were not disclosed throughout the article. A site-visit schedule was created for data collection purposes, and each firm was visited on a separate weekday. The data were collected based on a pre-determined schedule involving different time intervals of regular business hours.
Numerous workstations within a furniture manufacturing plant are designed to perform various production activities, such as wood and panel cutting, edge banding, sanding, drilling, surface finishing, upholstery, assembly, and packaging. In addition to those shop-floor components, there would also be warehouses and administrative offices as essential parts of any production facility. Some critical physical and chemical hazards and ergonomic risks exist in such a work environment. These risks include but are not limited to dust and noise exposure, VOCs, improper lighting, lack of proper air circulation, heavy lifting, trip hazards, or the use of tools with significant vibration.
Fig. 1. Illustration of the sampling procedure followed in the study
Each workstation includes different characteristics as a function of varying machine configurations and job specifications, leading to various dust, VOC, CO2 emission, and noise exposure levels. Therefore, selecting measurement locations was of critical importance for this study. To accurately designate the measurement locations within each facility, firms’ process flow diagrams and facility layout plans were created in collaboration with professionals from the firms ahead of the site visits. Minitab Quality Companion was used to create process flow diagrams, while AutoCAD 2021 was used to draw facility layout plans. The researchers used these process flow diagrams and layout plans to discuss and select the measurement locations at each plant.
Three measurements from five pre-designated workstations (departments), namely, sanding/finishing, panel/part cutting, assembly/upholstery, administrative offices, and warehouses, were taken at each facility for ambient noise level and indoor air quality detection. In addition to shop floor functions of corporate buildings, as some other studies suggested, administrative offices, packaging areas, and warehouses were also included for a better understanding of the indoor air quality and noise hazard status of corporate facilities (Nezis et al. 2019; Strelyaeva et al. 2019; Mannan and Al-Ghamdi 2021).
Data were collected in a time-phased manner using 1-hour intervals for three repetitions, namely, 09:00-10:00, 11:00-12:00, and 14:00-15:00. The length of each measurement was ensured to be at least five minutes before being recorded as a valid data point. The sampling strategy was developed with an understanding that machinery in a furniture manufacturing setting does not operate continuously at full capacity; however, it is acknowledged that the noise levels during active machine operation are critical for assessing the risk of hearing damage. Therefore, time intervals and length of the sampling procedure were strategically designed for accurate assessment of the noise levels that workers might experience during their shifts, factoring in both the periods when machinery was actively processing materials and when it was idle between two consecutive parts at each station. As such, the noise level analysis took into consideration the variance in noise levels during different operational phases, including a comparative analysis of peak noise levels (LAmax) that occur during active processing periods. A total of 15 data points were collected for each parameter used for noise level detection. One hundred thirty-five data points were collected from nine facilities for each evaluation parameter. Noise level measurements were sampled according to BS EN 9612 and ISO 1997-2-2017 standards without interrupting regular work sequences using a PBX LXTI Class I sound level meter. For noise level measurement purposes, data for six key parameters were collected. LAeq (A-weighted equivalent continuous sound level) was calculated by using Eq. 1 with a 5 dB(A) exchange rate, which meant that when the noise level was increased by 5 dBA, the amount of time a person could be exposed to a certain noise level to receive the same dose was cut in half (Fink 2017; The Engineering ToolBox 2004).
LAeq = 10 log [(1/T) ∫ (pA / pref)2 dt ] (1)
In Eq. 1, LAeq is equivalent sound level (dB), T is time period (s), pA is sound pressure (Pa, N/m2), and pref is reference sound pressure (2×10-5 Pa, N/m2).
Other noise parameters measured were LAmin (instantaneous minimum sound level), and LAmax (instantaneous maximum sound level) with percentiles of LA10, LA50, and LA90. During the measurements, the sound level meter was set according to ISO criteria at least 150 cm clear from any potential barriers, such as walls and machinery in the designated measurement station. The device was also placed close to the active working area to capture noise levels realistically.
Indoor air quality measurements involved three particulate matter categories: total particles, PM 2.5 and PM10, formaldehyde emission (HCHO), and carbon dioxide (CO2) level. Temtop M2000 2nd Air Quality Monitor was employed for indoor air quality measurements, and data collection was carried out with utmost care to ensure consistency and control for human-factor-driven variability. The measurement ranges of the device for the parameters mentioned above were 0 to 999 µg/m3, 0 to 2 µg/m3, and 0 to 5000 ppm, whereas the resolutions for the same parameters were 0.1 μg/m³, 0.001 mg/m³, and 1 ppm, respectively. Following the aforementioned ISO criteria, the device was held near the workers’ noses and mouths to measure the designated workstation’s air quality. The designated workstations’ temperature and humidity levels were also measured and recorded to check for potential abnormalities. The same time intervals and sampling length as the noise level sampling procedure were also followed for the indoor air quality data collection phase. A total of 15 data points were collected for each parameter tracked for indoor air quality detection.
All data were digitally stored and processed in Microsoft Excel before transferring to Minitab 18 Statistical Analyses software for descriptive and inferential statistical analysis. Upon completion of descriptive statistics, the inferential statistical analysis, two-way ANOVA, and Tukey Pairwise Comparisons were carried out on the study parameters. The two-way ANOVA analysis used a stepwise regression procedure involving second-degree interaction terms of independent variables.
RESULTS AND DISCUSSION
Results of Descriptive Statistical Analyses
Both independent variables, organization size, and department, were checked against thirteen dependent variables (evaluation parameters). Sample size, mean, minimum, and maximum values, range and median values, and Q1 and Q3 values across each parameter were reported in Table 1 as part of descriptive statistical analyses.
Table 1. Results of Descriptive Statistical Analyses
The mean value for PM 2.5 was 73.2 µg/m3, with min, max, and range values of 10.8, 379.4, and 368.6 µg/m3, respectively. Minimum and maximum values across all measurements for the PM10 variable were 4 and 70714 µm, respectively, which yielded an extensive range value of 70,710 µm. Similarly, a wide range (57,175 counts/L) value was recorded for the total particle variable with minimum and maximum values of 2023 and 59198 counts/L, respectively. As another indicator of indoor air quality, CO2 levels for different organization sizes and departments were measured. A mean value of 487 ppm was calculated based on 135 data points. The highest concentration of CO2 was measured to be 1089 ppm, whereas the lowest concentration level was 313 ppm. Another critical indicator of indoor air quality for furniture manufacturers is HCHO levels within the work environment. As presented in Table 1, the mean HCHO value was around 0.236 µg/m3 with minimum and maximum values of 0.0010 and 3.44 µg/m3, respectively.
Within the scope of noise level indicators, the mean LAeq value across all data points was 71.1 dB(A). The minimum LAeq value was 41.9 dB(A), while the median and maximum values for this variable were 74.7 dB(A) and 93.8 dB(A), respectively. The mean, minimum, maximum, and range values for the LAmin variable were 62.2, 33.3, 64.0, and 85.0 dB(A), respectively. On the other hand, a mean value of 83.0 dB(A) was recorded for the LAmax category. The minimum, median, and maximum values for this evaluation category were 56.7, 84.0, and 120.6 dB(A), respectively, as shown in Table 1.
Results of Inferential Statistical Analyses
The results of the inferential statistics showed that both organization size and department were statistically significant at the 95% confidence level in the means of PM 2.5 particles with p-values of 0.003 and <0.0001, respectively. Within the general linear model of PM 2.5 versus organization size and department, the interaction term of the independent variables was not significant at the same confidence level and had a p-value of 0.074, as shown in Table 2.
Table 2. Results of Two-Way ANOVA Analysis for PM 2.5 Observations versus Organization Size and Department
Tukey groupings of the size sub-groups indicated that small and mid-sized organizations had similar PM 2.5 levels, and the mean values of these size categories were not statistically different. In contrast, the large-sized organizations’ subgroup had a much lower mean value and was grouped in a different category, as given in Fig. 2A. Based on the Tukey groupings of sub-groups of the department type variable, sanding/finishing department and administrative offices were placed in the same group, while panel/part cutting, assembly/upholstery and warehouse departments were grouped. However, some departments were not statistically differentiated from each other. For instance, administrative offices were not statistically different from panel/part-cutting departments, and assembly/upholstery departments were not statistically different from warehouses, as shown in Fig. 2B.
Fig. 2. Tukey groupings of the PM 2.5 sub-group means for organization size (A) and department (B)
According to the general linear model of HCHO versus organization size and department variables, both independent variables and their first-degree interaction term were statistically significant at the 95% confidence level with p-values of 0.002, <0.0001, and 0.003, respectively, as presented in Table 3.
Table 3. Results of Two-Way ANOVA Analysis for HCHO Observations versus Organization Size and Department.
As shown in Fig. 3A, within the sub-groups of the organization size variable, medium- and large-sized organizations were grouped in the same category with much lower mean values (0.1619 and 0.1377) than that (0.4097) of the small-sized organizations. Within the scope of Tukey groupings of HCHO mean values across different departments, all departments but the sanding/finishing department were grouped under the same category, as given in Fig. 3B.
Fig. 3. Tukey groupings of the HCHO sub-group means for organization size (A) and department (B)
Two-way ANOVA results of LAeq versus organization size and department type yielded interesting findings; mean LAeq values across all organization sizes were not statistically different at the 95% confidence level, with a p-value of 0.077. Department type was found to be statistically significant with a p-value of <0.0001. The first-degree interaction term of the independent variables was also not statistically significant even though it was kept in the model by the stepwise regression procedure, as given in Table 4.
Table 4. Results of Two-Way ANOVA Analysis for LAeq Observations versus Organization Size and Department.
When Tukey groupings of the LAeq means of sub-groups belonging to organization size and department type variables were compared, as illustrated in Fig. 4A and 4B, all organization sizes were under the same category, while department types were under three distinct categories. LAeq mean values of panel/part cutting and sanding/finishing departments shared the same group, while administrative offices and warehouse departments were placed in another group. As shown in Fig. 4B, the assembly/upholstery department was individually grouped into another category.
Fig. 4. Tukey groupings of the LAeq sub-group means for organization size (A) and department (B)
Linear models constructed for variance analysis of other parameters, namely, total particles, PM10, LAmin, LAmax, LA10, LA50, and CO2, did not have strong enough R-square and adjusted R-square values (<0.3) when checked against organization size and department variables. Therefore, no interpretation of inferential statistics was carried out for those parameters. PM 2.5, HCHO, and LAeq were considered critical parameters and more detailly evaluated and discussed based on spatial patterns.
Fig. 5. Spatial layout illustration of average PM 2.5 for organizations
Spatial Patterns of Critical Parameters
A unique pattern emerged when looking at the spatial distribution of critical parameters, namely PM 2.5, HCHO, and LAeq, within the large, medium, and small firms across different departments. Color-coded (darker colors mean higher levels of concentration or emission) critical pattern measurements were also illustrated on representative facility layout maps of small, medium, and large-sized organizations. These are presented in Figs. 5, 7, and 9.
Fig. 6. PM 2.5 sub-group means for organization size (A) and department (B)
The average PM 2.5 values of the subgroups followed a unique pattern with some interesting findings when checked for organization size and department variables, as illustrated in Fig. 5. PM 2.5 values were also checked against the World Health Organization (WHO) guidelines (WHO 2010). According to these guidelines, there was no difference between the hazardous nature of particulate matter from indoor and outdoor sources. WHO highlights that the annual average concentration of PM 2.5 should not exceed 5 µg/m3, while 24-hour average exposures should not exceed 15 µg/m3 for more than 3 to 4 days per year. Among the measurements carried out for this study, the lowest average level of PM 2.5 concentration, 48.5 µg/m3, was measured in large-scale organizations and increased by the decreasing organization size, as shown in Fig. 6A. The average PM 2.5 value for large-scale organizations was more than three times higher than the safety threshold of 15 µg/m3.