Drinking water quality assessment from water dispensers in an educational institution

Access to clean drinking water presents a considerable challenge, as 2.2 billion people currently do not have access to safely managed drinking water, and 4.2 billion, or 55% of the world's population, are without safely managed sanitation (UNESCO 2020). In many countries, drinking water quality monitoring programs have been established to prevent or reduce the risk of contracting water-related infections. There has been a marked increase in the consumption of drinking water derived from different sources such as water dispensers a popular source of water in public places, workplaces and educational institutions. Such increased popularity is the result of easy access to clean water, and the convenience of the location, driven by post-COVID-19 education lifestyle changes and drinking habits (Schillinger & Du Vall Knorr 2004; CDC 2014; UNESCO 2020). However, concerns were raised that water dispensers had the potential to cause waterborne outbreaks, particularly in sensitive and immunocompromised populations (Farhadkhani et al. 2014; Al Moosa et al. 2015). Dissolved organic compounds in drinking water are responsible for the growth of bacteria and the colorization of water surfaces. Thus, drinking water already contained in water dispensers was found to be more contaminated than water newly supplied to the dispensers.

In Thailand, drinking water provided by dispensers is required by law to be free from any pathogenic microorganism as well as chemical contaminations (Thai Industrial Standards Institute 2006). However, previous investigations found that 76.6% of examined dispensers were contaminated with fecal coliform microbes, Escherichia coli, fecal streptococci, Pseudomonas aeruginosa or Staphylococcus spp. and revealed a prevalence of E. coli and S. aureus in filtration-treated dispenser water of 54.17% and 16.67%, respectively (da Silva et al. 2008; Wibuloutai et al. 2019). In particular, levels of E. coli and S. aureus are frequently monitored as indicators of the hygienic status of drinking water (Jessen et al. 2013). The Thailand Department of Health had recently reported that an estimated 40% of vending machines did not provide drinking water whose quality met the standard requirements (Wibuloutai et al. 2019). Consequently, unwanted changes in the microbial quality of drinking water can have adverse effects on the distribution system and consumers (Prest et al. 2016).

Given the importance of drinking water safety, this study was conducted to assess the factors influencing drinking water quality in water dispensers of three type-brands that are currently in use at Walailak University, Southern Thailand. The type-brands investigated here were STANDARD WP8-5, STANDARD CT4-DW1 and HYUNDAI ROMEO UF. All three types of dispensers feature a continuous direct connection to the piping water supply, technical differences relate to the number of filtration processes and the materials involved. The dispensers also differ in the number of faucets. The dispensers Standard WP8-5 and Hyundai Romeo UF use stainless steel as tank material while the dispenser Standard CT4-DW1 has a cast aluminum tank. The purification process commonly involves reverse osmosis and carbon filtration, while UV radiation is rarely used in commercial stand-alone units. The Hyundai Romeo UF dispenser uses nano-pH filtration in addition to conventional ultrafiltration.

The objective of this study was to investigate the microbiological and physicochemical quality of drinking water plumbed-in water dispensers and to assess the sanitation practices in an educational institution in an area of Southern Thailand.

All water samples were collected from September to November 2020 in Walailak University (WU), the largest public university in Thailand, located in Tha Sala District, Nakhon Si Thammarat province, at 3.030N and 101.70E. A total of 120 drinking water samples were collected in sterile polyethylene bottles from 60 water dispensers located in four main zones in WU, consisting of a number of samples from each zone, including 1. academic building zone (n = 35), 2. lecture building zone (n = 35), 3. dormitory zone (n = 40), and 4. canteen/cafeteria zone (n = 10). The three different type-brands of water dispensers used in all study areas were STANDARD WP8-5, STANDARD CT4-DW1 and HYUNDAI ROMEO UF (Figure 1).

The drinking water samples were analyzed for pH, turbidity and hardness by electrometric measurement, nephelometry and EDTA titration, respectively (Public Health Act 1992). Total coliform bacteria and fecal coliform bacteria were quantified by Most Probable Number (MPN) test. In addition, the microbiological analysis investigated for the presence of Escherichia coli (E. coli) and coliform bacteria using membrane filtration with 3M Petrifilm^TM for E. coli/coliform count (EC) (Bird et al. 2020). All analytical methods were performed according to the standard methods for the examination of water and wastewater (Public Health Act 1992; Eugene et al. 2012). Residual free chlorine in water was determined by using a commercial test kit (Haier Enterprise, Pathum Thani, Thailand).

A questionnaire for the survey of the environmental conditions around each water dispenser at the time of sampling was designed using a modified evaluation form (Public Health Act 1992; Handbook of Drinking Water Vending Machines 2013).

Statistical analysis was performed using IBM SPSS Version 22. Descriptive analysis was used to describe measures of tendencies and dispersion of the water dispenser quality. The Chi square test was used for the assessment of factors influencing the drinking water quality in the water dispensers. Statistical significance was assessed using two-sided tests with a p-value of 0.05.

The study was reviewed and approved by The Human Research Ethics Committee of Walailak University, approval number WUEC-20-153-01.

All water samples from water dispensers were analyzed by both physiochemical and microbiological procedures and the summary of the results is shown in Table 1. Samples for each water dispenser were collected in duplicate and the mean of laboratory and surveillance results was used for the statistical analysis. It should be noted that differences between the samples were relatively small as expected, since both samples were collected at the same time point. For all samples, the pH, color, turbidity, total solids, TDS and free chlorine did not exceed the reference values of the drinking water regulations, whereas total hardness (CaCO₃) of water samples ranged from 2.79 to 124 mg/L and 13.33% of the samples had slightly higher concentrations than the reference limit (APHA 2012). The microbiological results indicated that coliform and fecal coliform bacteria were found in 16.67% and 8.33% of the samples, respectively. Escherichia coli was not detected in any of the water samples. It should be noted that the testing methodology for total coliform bacteria reports numeric values as MPN/100 mL, while the results for E. coli are given as CFU/100 mL, which can potentially lead to an underestimation of the E. coli count. The concentration of residual free chlorine in the samples investigated was below the minimum recommended by WHO, thus suggesting the likely presence of disease-causing microorganisms (Table 1). The vending machines evaluated in this study do not provide chlorination as they are fed with treated water from the mains supply, which ideally should be coliform free.

There were 60 water dispensers in operation, which were evaluated for sanitation and the surrounding environmental conditions according to six items (and 22 sub-items) as listed in the Handbook of Drinking Water Vending Machines and the Ministry of Public Health drinking water standard, 1992. For the assessment of factors influencing drinking water quality, the relationship between the sanitary parameters of six factors: (1) location and pathogen source around water dispensers (5 sub-items), (2) dispenser conditions (4 sub-items), (3) water resource and quality improvement (3 sub-items), (4) drinking water standardized quality control (2 sub-items), (5) maintenance and service (5 sub-items), and (6) water quality records and monitoring (3 sub-items), were assessed.

The criteria for assessing the sanitization and the water dispenser surroundings were evaluated by a questionnaire and comprised six items (and 22 sub-items). Scores were allocated to the examined dispensers by using 3-point scales (Jacoby & Matell 1971) and ranged from 1 to 3 (‘strongly disagree’ to ‘strongly agree’). The mean score was categorized into 3 groups as follows: 1.00–1.49 scores as covering none of the items listed as ‘should-improve’ level; 1.50–2.49 scores as covering some items as ‘poor level’ and 2.50–3.00 scores as covering all items as ‘good level’.

The results of the evaluation are shown in Table 2. We have evaluated and assigned scores to six main items and 22 sub-items to identify possible correlations of statistical significance. We have then analyzed by using chi square test which of the variables have the highest impact on data association, which were (with a calculated statistically significant p-value of <0.05): i) location and pathogen source of the water dispensers, ii) hygienic conditions and iii) maintenance and service, indicating that only three out of six evaluated items exhibited a significant correlation to performance of the machine. The results also revealed that type-brand STANDARD WP8-5 met the highest water quality standards, followed by the models HYUNDAI ROMEO UF and STANDARD CT4-DW1, respectively (Figure 2).

With the objective of evaluating the drinking water quality provided by public water dispensers, we have analyzed physicochemical and microbiological parameters of water samples collected at Walailak University, an institution of higher education in Southern Thailand. In particular, we have focused on a comparative analysis of the bacteriological quality of drinking water provided by commercial dispensers.

Pollution of drinking water caused by fecal contamination causes concerns due to the risk of contracting diseases by microbial pathogens. Fecal coliform bacteria indicate contamination of water with fecal waste that may contain other pathogenic organisms, including bacteria, viruses, or parasites such as Cryptosporidium and Giardia, the causative agent of beaver fever (Srisuphanunt et al. 2010). In analogy to previously published reports, total coliform bacteria, fecal coliform bacteria and E. coli were enumerated (Schillinger & Du Vall Knorr 2004; da Silva et al. 2008). Ten out of 60 samples tested for the presence of total coliform bacteria (16.67%) and five out of 60 samples tested for fecal coliform bacteria (8.33%) were found to be unsatisfactory concerning microbiological quality. E. coli could not be detected in any sample, a finding in accordance with similar studies recently conducted (Pratum & Khananthai 2017; Yongyod 2018; Wibuloutai et al. 2019). The absence of these microorganisms, considered to represent an indicator of fecal contamination, renders water quality satisfactory and safe without health concerns. A recent study conducted in Malaysia observed unacceptable levels of E. coli bacteria in water samples from Seri Serdang and Taman Pinggiran Putra that exceeded the WHO recommendations (Tan et al. 2016). However, it should be noted that the species of bacteria were not identified in our study, rendering the possibility that pathogenic bacteria such as fecal streptococci, Pseudomonas aeroginosa and Staphylococcus spec. are still contained in the samples. Recent studies have proposed performing a periodic enumeration of Pseudomonas to maintain water quality in municipal water systems (da Silva et al. 2008). Therefore, the microbial content in our samples would warrant further investigations into possible health effects on water consumers.

The physiochemical parameters of the samples were found to be within safe limits issued by WHO. Temperature, pH and dissolved organic compounds can influence the growth of bacteria on the surfaces of water dispensers (Gibert et al. 2013). Eight out of 60 samples were found to exceed acceptable limits for total hardness (as CaCO₃) (Thai Industrial Standards 2006). Water hardness has always been considered as an important factor causing clogging of membrane filters (Handbook of Drinking Water Vending Machines 2013). Besides, leakage of pipelines, improper disinfection by chlorination, biofilm development and improper performance of the filtration unit may be reasons for this complication. It should be considered, however, that water hardness is a factor of minor relevance in Thailand, as the surface water is mainly sourced from rain.

Our findings suggest that machine maintenance and regular servicing intervals are important factors for the prevention of microbial contaminations. A significant proportion of the dispensers surveyed in our study had no regular maintenance records, which resulted in ‘should-improve’ levels of 42/60 (70%) for maintenance and service, and 45/60 (75%) of water quality records and monitoring, respectively. Recent reports have shown that contamination with fecal bacteria may not only occur due to unsanitary hygiene practices of people handling the dispenser machines but also through the attrition of pipes and corrosion. These problems may be contributing factors not only to bacterial contamination but may also lead to contamination by toxic chemicals (Ebrahim et al. 2015; Mumbi & Watanabe 2020). Leakage and cross-contamination between sewage pipes and drinking water pipelines or due to leaky and worn out pipes may be one reason for this contamination.

Consequently, our study recommends regular maintenance and cleaning to prevent microbial contamination associated with the faucets and water dispensers.

This study presents information regarding water quality and technical conditions for water dispensers in an educational institute. We show that coliform bacteria and fecal coliform bacteria are major contaminants in drinking water that are effective, but not completely controlled by the machines. Regular maintenance and cleaning of water dispensers would greatly diminish the chances of contamination of this water source and reduce the risk to human health. This research may also enable concerned authorities to set up effective programs and remedial actions to address prevalent problems regarding drinking water quality in educational institutions.

This research project was supported by School of Allied Health Sciences, Walailak University. We would like to thank Miss Aporn Chaisuwan, Librarian, Walailak University, Thailand, for assistance with the references.

The authors declare no conflicts of interest.

This work was supported by the School of Allied Health Sciences, Walailak University.

All relevant data are included in the paper or its Supplementary Information.