Comparison of sampling collection strategies for assessing airborne trichloramine levels in indoor swimming pools

Since 1995, Hery’s trichloramine sampling procedure has been widely used to determine trichloramine exposure in indoor swimming pools. This method consists of pumping air at a 1 L/min flow rate for 2 h through a Teflon prefilter and two quartz fiber filters. Modified Hery methods have been reported using different sampling pump flow rates and types of prefilters. It is possible that the prefilter type or sample collection pump flow rate influenced the results of these studies. This study is designed to evaluate the effects of different cassette assemblies and sampling flow rates on the levels of measured trichloramine. Laboratory tests were performed using a trichloramine production setup designed for this study. Workplace measurements were carried out at four indoor swimming pools. Different prefiltering strategies were used: no prefilter, glass prefilter or Teflon prefilter in the sampling cassette, and an original separable prefilter cassette is presented in this study. Laboratory tests indicated that at trichloramine concentrations higher than 1 mg/m3, the percentage of trichloramine captured on the first filter could be less than 75%, which demonstrated possible loss of the material during sampling. An investigation of the prefilter effect on the sampling strategy using different cassette assemblies revealed that using a separable cassette assembly prevented overestimations of trichloramine levels. Furthermore, there were no significant differences between trichloramine concentrations measured at flow rates (from 0.5 to 2 L/min) in swimming pools.


Introduction
Swimming pool water is subjected to disinfection to inactivate microbial pathogens and prevent waterborne diseases (Ilyas et al. 2018). Among the four main classes of disinfectants (chlorine-based, bromine-based, ozone, and UV), chlorine is most commonly used for indoor swimming pools (Teo et al. 2015;Wyczarska-Kokot et al. 2020). However, unintended reactions between chlorine and the natural organic matter present in water (nitrogen compounds brought by bathers) lead to formation of disinfection byproducts, including chloramines (Chowdhury et al. 2016;Reckhow et al. 1990;Teo et al. 2016). Several recent studies have characterized exposure to disinfection by products in water and air using measurement or modeling approaches (Kumari and Gupta 2015;Mahato and Gupta 2021;Mahato and Gupta 2022;Peng et al. 2020). However, trichloramine (TCA) has the highest concentration among other volatiles in indoor swimming pool air (Ahmadpour et al. 2022;Weng et al. 2011;Yang et al. 2018). More specifically, due to the high volatility (vapor pressure = 19.95 kPa at 20 °C) and relatively low water solubility of trichloramine compared to monochloramine and dichloramines, trichloramine is mainly dispersed in the gas phase (Manasfi et al. 2017;Ratnayaka et al. 2009).
Numerous reports of health issues showed that a high prevalence of airway irritation in a swimming pool facility could be explained by high TCA exposure levels (Fantuzzi et al. 2012;Levesque et al. 2015;Mahato and Gupta 2022;Parrat et al. 2012;Westerlund et al. 2022). In this context, Hery et al. (1995) were the first to find an association between airborne TCA levels and acute respiratory symptoms among instructors at swimming pools (Hery et al. 1995). Further studies reported that TCA caused ocular and respiratory symptoms in swimming pool employees (Chen et al. 2008;Fantuzzi et al. 2010;Jacobs et al. 2007;Massin et al. 1998;Nordberg et al. 2012). In a study of 41 indoor swimming pools in Quebec (Canada), Tardif et al. (2016) reported a median TCA concentration of 0.23 mg/m 3 , and highest concentration was 0.56 mg/m 3 (Tardif et al. 2016).
Other studies reported concentrations of TCA ranging from 0.1 to 2.2 mg/m 3 (Carter and Joll 2017;Manasfi et al. 2017;Richardson et al. 2010). For workers' protection, the German Federal Environmental Agency proposed airborne TCA levels of 0.2 mg/m 3 (German Working Group on Indoor Guide Values of the Federal Environment Agency 2011), while the French Agency for Food, Environmental and Occupational Health & Safety (ANSES) recommended a maximum of 0.3 mg/m 3 for an 8-h exposure (ANSES 2010). Among Canadian provinces, only British Columbia set an exposure limit concentration, and this was 0.35 mg/m 3 (Westerlund et al. 2019) for an 8-h exposure limit.
No standard sampling and analytical methods have been published for TCA by mandating organizations such as OSHA (Occupational Safety and Health Administration), US EPA (United States Environmental Protection Agency), and NIOSH (National Institute for Occupational Safety and Health). However, data reported from environmental sampling during 2 h with Hery's method have been used to determine occupational exposures (Ahmadpour et al. 2022;Saleem et al. 2019;Westerlund et al. 2019). Hery's analytical and sampling method consists of pumping air at a 1 L/min flow rate for 2 h through a sampling cassette made up of a Teflon prefilter (polytetrafluoroethylene) and two quartz fiber filters (QFFs) impregnated with sodium carbonate and arsenic trioxide. In the presence of sodium bicarbonate, chloramines are decomposed into ammonia and hypochlorite, and in the presence of arsenic trioxide, hypochlorite ions are then reduced to chlorides. These chlorides are then quantified by ion chromatography. Any other chloride (mono and dichloramine) that could be present in the air (in the form of droplets) should be eliminated by the prefilter, while TCA passes through the prefilter and is then trapped on the QFFs. However, Hery et al. (1995) indicated that transfer of chlorides from the Teflon prefilter to the QFFs could potentially occur due to quick vaporization and adsorption (Hery et al. 1995), which is known as migration. TCA vapors will pass into the primary QFF and bind to it. When the QFF becomes saturated, contaminants will then pass into the second QFF (called the back-up section), which is known as breakthrough.
Environmental parameters such as temperature and humidity can influence the adsorption of gases and vapors on substrates. According to OSHA, high temperatures can contribute to the migration of contaminants in the two sections of charcoal tubes used to sample VOCs (OSHA 2014). In addition, when sampling is performed in an environment with high humidity, water vapors complete for the binding sites of the sorbent. OSHA recommends modifying sampling parameters (flow rate and sample duration or air volume) to take these environmental conditions into account. Other OSHA methods recommend refrigerating samples or using separable sorbents to avoid migration of the contaminants. Indoor swimming pools are sites high temperatures and high humidity.
The past thirty years have seen rapid advances in TCA exposure assessment, and several studies have assessed TCA levels in indoor swimming pools using different sampling strategies (Soltermann et al. 2014;Wu et al. 2021;Zwiener and Schmalz 2015). Modified Hery methods have been reported with different sampling pump flow rates and types of prefilters. While some studies did not use prefilters (Massin et al. 1998;Person et al. 2005;Westerlund 2016;Westerlund et al. 2019), others replaced the Teflon prefilter with glass fiber filters (Levesque et al. 2015). In addition, different flow rates ranging from 0.25 (Westerlund 2016) to 4 L/min (Person et al. 2005) were used for sampling. For instance, some authors have used low flow rates to avoid rapid passage of TCA through the QFFs (Westerlund 2016). The sampling parameters (e.g., prefilter and flow rates) used in different published studies are reported in Table 1.
Whether due to migration or breakthrough, inadequate sampling conditions could result in a loss of TCA adsorbed on the QFFs and consequently an underestimation of TCA concentrations (Drolet andBeauchamp 2012, NIOSH 2017). To the best of our knowledge, no study has assessed the impact of modifying sampling parameters (using a prefilter or not, changing the flow rate) on the TCA concentrations determined with an experimental setup and field sampling.
This study presents a comparison of different sample collection approaches for measuring TCA in the air of indoor swimming pools. More specifically, this study is designed to assess the effects of TCA concentrations, the prefilter (cassette assembly), and the sampling flow rate on TCA measurements.

Methods
TCA sample collection methodologies were evaluated using two different test series: laboratory tests and field experiments at swimming pools. As shown in Table 1, a limited number of studies (only one published) used glass as a prefilter, and the majority used Teflon. Additionally, most studies took samples with a sampling pump flow rate from 0.5 to 1.25 L/min. Only one publication used 0.25 L/ min, and another used 4 L/min. Although we performed tests for these minority cases in our study, we tested the most common prefilter and sampling pump flow rate extensively.

Cassette assemblies
Impregnated filters were prepared based on the Hery method by using 37-mm QFFs (Pall fabric, pure quartz, no binder, 2500 QAT-UP, 37 mm). These filters were deposited inside 37-mm closed-face cassettes (1/4-inch inlet diameter). The QFFs (without any pretreatment) were soaked with a solution of arsenic trioxide and sodium carbonate. The impregnation solution was prepared by dissolving 0.8 g As 2 O 3 and 5.0 g Na 2 CO 3 in 100 mL of Milli-Q water. Using these filters, five series of cassettes (Supplementary Information- Figure S2) were prepared: (1) (TQQ), one cassette with three layers: in the first layer, Teflon was placed as a prefilter, and in the second and third layers, two impregnated QFFs were placed (as described in Hery's method).
(2) (GQQ), one cassette with three layers: in the first layer, a glass filter was placed as a prefilter, and in the second and third layers, two impregnated QFFs were placed. (3) (QQ), one cassette with two impregnated QFFs and without a prefilter. (4) (T + QQ), two cassettes, with the first Teflon (T) prefilter connected via a Tygon tube to the second (QQ) cassette ( Figure S3).
(5) (G + QQ), two cassettes, with the first a glass fiber (G) prefilter connected via a Tygon tube to the second (QQ) cassette.
Cassettes were sealed with a cellulose shrink band. For the T + QQ and G + QQ cassette assemblies, prefilter cassettes T or G were connected to the QQ cassettes via a 20-cm Tygon tube (inner diameter = 1/4 inch). The two parts were disconnected immediately after sampling. The sampling cassettes were connected to a sampling pump with a 90-cm Tygon tube. TCA was sampled using Gilian Gilair5 pumps (Sensidyne, Clearwater, FL, USA). The cassettes were kept at 4 °C with the ends closed and sealed before analysis by ion chromatography. All samples were analyzed within one week after sampling.

Trichloramine analytical method
After removing the QFFs, each filter was placed individually in a 15-mL conical centrifuge tube (two tubes per cassette). The contents of each tube were dissolved in 10 mL of Milli-Q water. The mixture was mixed for 30 s, and TCA extraction was continued at room temperature for 10 min in an ultrasonic bath (Branson, model 3800). Next, 5 mL of the filtered solution was transferred directly into vials for analysis using with ion chromatography-conductivity detection (IC-CD) (Thermo Scientific DionexTM Aquion™-Guard Column: Dionex IonPac AG22 (4 × 50 mm). Column Type: Dionex IonPac AS22 (4 × 250 mm)). The eluent, sodium carbonate/bicarbonate concentrate (Thermo Fisher-Dionex), was used as a mobile phase (1.2 mL/min). To improve the sensitivity of the analytical method and the linearity of the calibration curve, a suppressor (Dionex AERS 500 carbonate) was added, and a current of approximately 33 mA was applied. Six control samples with chloride concentrations ranging from 0.2 to 20 mg/L and at least two blanks per cassette type were run along with the samples. Chloride concentrations detected in the blanks were subtracted from the concentrations in the samples. Each filter required a 20-min run time for ion chromatography. The method detection limit was 0.09 mg/L chloride, which corresponds to the lowest detectable concentration of 0.011 mg/m 3 TCA for 2-h sampling at a flow rate of 0.75 L/min (90 L of air). TCA air samples were analyzed in the Department of Occupational and Environmental Health Laboratory at the Université de Montréal, Québec, Canada.

Laboratory study design
A TCA production setup was assembled by following the description of Hery et al. (1998) (Hery et al. 1998). The trichloramine generation setup had three sections: (1) chemical combination, (2) processing, and (3) dilution. Figure 1 shows a diagram of the setup. The original photo is provided in supplementary information ( Figure S1). Production of TCA was conducted in a continuously stirred tank reactor (2-L three-neck flask), which contained an initial sodium hypochlorite (Fisher Chemical™ 5.65-6%/ Laboratory) concentration of 3 to 3.5 mmol/L, and an initial ammonium sulfate (Fisher Chemical™, ACS grade) concentration of 0.5 to 0.7 mmol in a phosphate buffer solution (900 mL). Indeed, all tests were completed with a pH between 3 and 4 and a hypochlorite to ammonium sulfate ratio around 10 to promote the formation of TCA. An ammonium sulfate solution (1 mM) and a sodium hypochlorite solution (10 mM) were continuously added to the reactor at equal rates of 0.7 mL/min with peristaltic pumps.
Nitrogen gas was added directly to the reactor flask to transfer the chloramines from the liquid phase to the scrubbing reactor in the gas phase while aiding in stirring and mixing the system. Compressed air was added to the system in the dilution section, which was used to dilute TCA vapors and bring them to the desired concentration range. For the processing section, we used two scrubber reactors; the first one was filled with 500 mL of sulfamic acid (20 mmol, ACROS, 99%) to trap any free chlorine present, and the second scrubber contained Milli-Q water to trap mono-and dichloramines. In the dilution section, the airflow containing TCA entered the dilution box, and the compressed air was added to the system to dilute the TCA vapors (100 to 200 dilution factor). The dilution rate was calculated to produce a concentration of TCA within the range of concentrations reported in the literature (1 to 1.92 mg/m 3 ). There was no complementary method for analyzing the TCA concentrations before collecting the samples. However, the diluted TCA vapors entered a parallel sampling line allowing up to four side-by-side samplings ( Figure S4a). A Tygon tube (Versilon™ C-219-A) was used to connect the cassettes to the pumps.
To study the effects of TCA concentrations on filter saturation, 44 tests (11 series) with a wide range of TCA concentrations were run. The QQ cassette assembly with a 1 L/min pump flow rate was used for 2 h. Six samples were lost in the Fig. 1 Schematic of the TCA production setup (1: ammonium sulfate, 2: sodium hypochlorite, 3: phosphate buffer, 4: nitrogen gas, 5: pH meter, 6: scrubbing reactor containing sulfamic acid, 7: scrubbing reactor containing Milli-Q water, 8: peristaltic pump, 9: air, 10: mixing chamber, 11: TCA sampling) analytical procedure, and the results from 38 samples were considered. The cassette type and sampling pump flow rate were kept constant for the laboratory samples. The percentage of TCA captured in the first QFF was considered an indicator of TCA sample collection efficiency. The target TCA concentrations in this study ranged from 0.1 to 2 mg/m 3 .

Swimming pool sampling design (field experiments)
Four public indoor swimming pool facilities in three cities of Quebec (Canada) were included in our study. The facility hall volumes ranged from 2890 to 13,556 m 3 , and the number of visitors per year varied between 29,000 and 40,000. The main characteristics of each facility are presented in the Supplementary Information (Table S1). The swimming pools selected were known to have high TCA concentrations in the water and air or a high number of complaints about environmental conditions from swimmers/workers. In all facilities, the sampling sites were located 50 cm above the water surface and 0-1 m from the pool's periphery. Several studies used a similar methodology in their researches where samples were collected at 30 cm to 150 cm above the water surface (Catto et al. 2012;Tardif et al. 2016). Six days of sampling were performed (one-day visits at swimming pools 1 and 2 and two-day visits at swimming pools 3 and 4) to study the flow rate effect (from 0.25 to 2 L/min, sample volumes 30-240 L) and the cassette assembly effect (QQ, TQQ, GQQ, G + QQ, and T + QQ cassettes) on the TCA concentration measurements. The sampling time in all swimming pools was two hours.

Data analysis
All data analyses were performed with Microsoft Excel (V2019, Microsoft) and Statistical Package SPSS, version 27 (IBM Inc., Chicago, IL, USA). For each cassette assembly and sampling flow rate, TCA concentrations were calculated by adding the two QFF measurements. The second filter played the role of a backup filter, and the TCA% captured on the first QFFs was calculated. To the best of our knowledge, there is no standard breakthrough criterion published for samples taken by QFFs. However, a 25% cutoff value is indicated to determine the validity of the solid sorbent tube applied for collecting organic vapors and gas samples (McCammon and Woebkenberg 1984). Applying the 25% criterion to impregnated QFFs is thus considered a conservative approach and a common practice for measuring occupational exposure. For descriptive purposes, the arithmetic mean (AM), standard deviation (SD), and median were calculated to describe TCA concentrations. Data were analyzed using a one-way analysis of variance (ANOVA) to compare means for TCA% captured on the first QFF with different sampling pump flow rates and prefilter assemblies.
Hery's TQQ method was used as the standard comparison for the ratio calculation. The ratios were calculated by dividing the TCA concentration obtained with the separable cassette assembly (T + QQ and G + QQ) or without a prefilter (QQ) by the concentration obtained with the corresponding sample collected using the TQQ cassette assembly from all of the swimming pools. Paired t test analyses were carried out between ratios with separable and inseparable cassettes. The level of significance for these analyses was set at 5%.

Laboratory sampling
TCA concentrations tested ranged from 0.12 to 2.07 mg/m 3 . The TCA% captured on the first filter of the QQ cassettes ranged from 45 to 100%. Figure 2 presents the distribution of TCA% on the first filter as a function of the TCA concentration. These results showed that when the concentration of TCA was lower than 1 mg/m 3 , breakthrough was less probable since more than 75% of the TCA was captured by the first QFF. Table 2 presents an overview of the sampling and study design as well as the descriptive TCA results for the six sampling days. The mean TCA concentrations in swimming pools 1 and 2 were 0.14 and 0.77 mg/m 3 , respectively. For swimming pools 3 and 4, the mean TCA concentrations were 0.90 mg/m 3 and 0.83 mg/m 3 for the first visit and 0.45 mg/ m 3 and 0.36 mg/m 3 for the second visit, respectively. Figure 3 presents the TCA% captured on the first QFFs according to the cassette assemblies. The values for TCA% captured on the first QFFs (AM% ± SD%) were 91% ± 1% (G + QQ), 87% ± 9% (GQQ), 88% ± 8% (QQ), 90% ± 3% (T + QQ), and 91% ± 4% (TQQ). For all the tests, the TCA% on the first QFF was greater than 75%, and no significant variations were observed among cassette assemblies (p < 0.05). Figure 4 provides TCA concentrations as a function of cassette assembly for each sampling pool/visit. TCA concentrations determined for swimming pools by using different sampling cassette assemblies did not demonstrate significant variation. The highest variation was approximately 14% for sampling with a separable Teflon prefilter (T + QQ) and sampling without a prefilter (QQ) in swimming pool 2.

Effect of cassette assemblies
Significant differences were observed between samples using a separable prefilter and an inseparable prefilter (p < 0.05) (Fig. 5). The AM ± SD values of the ratios T + QQ/ TQQ and QQ/TQQ were 0.93 ± 0.06 and 0.99 ± 0.13, respectively. This finding suggested that using a separable prefilter could affect the measured concentrations. An average ratio of approximately 1 (0.99 ± 0.13) for QQ/TQQ indicates no difference between sampling with a prefilter in the same cassette (inseparable cassette assembly) and sampling without a prefilter. All these results together seem to support the possibility of chloride transfer from the Teflon prefilter to the QFFs.

Effect of sampling pump flow rates
For the 84 tests carried out in the indoor swimming pools, the TCA concentrations were between 0.12 mg/m 3 and 1.00 mg/m 3 . The cassette assembly details and TCA concentrations are presented in Table 2. The TCA% captured on the first QFFs were greater than 75% for all these samples, and no variation was observed for different sampling flow rates (Fig. 6). Figure 7 presents the TCA concentrations measured during two different visits at swimming pools 3 and 4. The results of samples taken with the TQQ cassette assembly from swimming pool 3 are presented in Fig. 7a and b. Comparing the medians of samples with different flow rates during the two visits did not indicate significant differences. Figure 7c and d presents data for the samples taken with the T + QQ separable cassettes during two   Fig. 3 Swimming pool sampling results. Effects of different sampling cassette assemblies on the TCA% captured on the first QFF. Samples were collected from swimming pools for 2 h at a 1 L/min flow rate. TCA concentrations ranged from 0.12 to 0.92 mg/m. 3 different visits to swimming pool 4. No significant differences were observed upon sampling with different flow rates. These results match those observed for swimming pool 3. However, significant daily variations (p < 0.05) were measured between the two sampling days in the two swimming pools, with average concentrations of 0.90 ± 0.07 and 0.45 ± 0.02 mg/m 3 for Day 1 and Day 2 in swimming pool 3 and 0.83 ± 0.04 and 0.36 ± 0.07 mg/ m 3 for Day 1 and Day 2 in swimming pool 4, respectively.

Evaluation of the effect of TCA concentration on the TCA% captured on the first QFF
One objective of this paper was to study the effect of TCA concentration on the TCA% captured on the first QFF so as to validate the method. Our laboratory test results showed that high TCA concentrations (TCA > 1 mg/m 3 ) could lead to a high TCA% on the second filter, which indicated possible breakthrough. Consequently, these results show that the results of these air samples could have been underestimated due to sample loss. Although most studies have reported TCA air concentrations in indoor swimming pools lower than 1 mg/m 3 , some pointed out swimming pools with higher concentrations. For instance, a TCA concentration of 1.06 mg/ m 3 was reported for a water park (Stansbury et al. 2009) and 1.92 mg/m 3 for the bubbling bath atmosphere of a recreational pool (Hery et al. 1995). These authors did not report the values of TCA% captured on the first QFFs. Further work is required to validate the TCA sampling methodology for swimming pools with concentrations higher than 1 mg/m 3 .

Evaluation of the effect of prefilter on the TCA % captured on the first QFFs and the level of measured TCA
The second part of this study was designed to assess the importance of prefilters in TCA exposure measurements. For sampling with TCA levels in swimming pools ranging from 0.12 to 1.00 mg/m 3 , the results revealed that cassette assembly was not associated with a high TCA% in the second filter and that these cassette assemblies protected the filter against breakthrough (TCA% on the first QFFs > 75%). TCA concentration ratio (R/TQQ) for a separable cassette assembly and sampling without a prefilter in swimming pool tests (number of tests = 42). The whisker plot presents the range and median of the TCA concentration ratio. The median TCA concentration for samples measured without a prefilter was higher than that for samples taken with the Hery TQQ cassette assembly, and this ratio was lower in samples with separable cassette assemblies The effect of cassette assembly on the level of measured TCA was evaluated by parallel samplings ( Figure S4). Significantly lower concentrations were measured when using a separable prefilter (p < 0.05) (Fig. 5). Using the separable prefilter seemed to prevent the migration of chlorides other than TCA onto the QFFs.
However, when the prefilter was placed inside the cassette (TQQ), no differences were seen compared to sampling without the prefilter (QQ). This finding is in accordance with previous studies showing no differences with or without Teflon prefilters used for sampling (Schmalz et al. 2011;Westerlund et al. 2015). This may indicate migration of contaminants inside the cassette (from the prefilter to the QFFs). Overall, this finding supports the hypothesis formulated by Hery regarding vaporization and transfer of chlorides other than TCA from the prefilter to the first QFF (Hery et al. 1995) and suggests that using a separable prefilter could Fig. 6 Effect of sampling pump flow rates on TCA% captured on the first QFF. Samples were collected from swimming pools for 2 h with a TQQ cassette assembly Fig. 7 Flow rates for sampling total TCA concentrations during visits to swimming pools number 3 (a, b) and 4 (c, d).
(a, b) The results of two visits to swimming pool 3. (c, d) The results of two visits to swimming pool 4. Sampling in swimming pools 3 and 4 was conducted with TQQ and T + QQ cassette assemblies, respectively. The dashed line shows the median TCA concentration improve air chloride separation and limit the associated uncertainty. To better understand the importance of prefilters or separable filters in TCA measurements, studies should be performed in pools that generate considerable quantities of droplets in the air.
Evaluation of the effect of sampling pump flow rate on the TCA% captured on the first QFF With the concentrations and sampling times assessed in this study, no sufficient changes were observed that could call into question the validity of breakthrough protection. The sampling results for swimming pools with flow rates of 0.5 to 2 L/min indicated that the first filter concentration was systematically higher than the cutoff value (TCA% > 75%).
Different authors have measured TCA levels by using a range of sampling flow rates from 0.25 to 4 L/min, and they reported TCA concentrations from 0.1 to 1.92 mg/m 3 (Table 1). In our study, the maximum flow rate studied was 2 L/min. Hence, sampling with a high flow rate in highly exposed swimming pools will increase the risk of underestimation due to breakthrough. Professional judgment is thus necessary in these situations.

Conclusions
To the best of our knowledge, this is the first study aimed at evaluating different sampling strategies used to monitor TCA concentrations in indoor swimming pools.
Our laboratory experiments showed that at TCA concentrations higher than 1 mg/m 3 , more than 25% of the collected TCA was deposited on the second filter, indicating the possibility of loss and underestimation of TCA levels. Our results support the use of a separable cassette assembly (i.e., T + QQ) to address the migration problem. Furthermore, in the swimming pool studies, TCA concentrations measured by different methods were quite similar for each sampling pool per day. This study highlighted the validity of Hery's adapted methods. The results of this research help future studies to select the sample collecting flow rate from the range of 0.25 to 2 L/min based on their preferences without questioning the possibilities of overestimation or underestimation of TCA levels. The outcomes of two campaigns in the same pool showed that the daily changes are more important than the differences related to different sampling methods. Further research should be focused on characterizing environmental changes in indoor swimming pools.
The present study provides useful information on TCA levels in swimming pools in Quebec. The findings are also valuable for understanding TCA sample collection methods and their comparability.