Water usage pattern of peri-urban community in the river basin: analyzing crucial variables for human exposure model

The human exposure model (HEM) is an important tool to predict the amount of contaminants carried by water to the community. Proper input variables play a crucial role in running the HEM, which is often scarce, particularly in low- and middle-income countries like Indonesia. To overcome that problem, we conducted a water demand survey of 10 water-related activities in a peri-urban community of the Upper Citarum River Basin. The survey resulted in three main variables of the exposure model from those activities, i.e. frequency, volume, and duration. This study aims to analyze if the contaminant uptakes from the HEM output were significantly different between groups (i.e.: male and female groups, 18 – 65 y.o. and ≥ 65 y.o. groups) and which exposure routes and water-related activities contribute the most to the contaminant’s total uptake by using statistical and proportional analysis. The output of carbofuran deterministic modelling showed the average total uptake was 1.06.E+00 ng/kg/day. The statistical analysis revealed a significant difference between oral and dermal uptake, but there was no significant difference in total uptake between gender and age groups. After conducting the proportional analysis, it was found that oral uptake contributed almost 99% to the total uptake of carbofuran. The analysis also showed that drinking and cooking were two activities contributing the most to the total uptake. This study’s results are necessary for health risk assessment and HEM development, especially in the field of river basin management.


Introduction
Water insecurity has been a global concern and significantly affecting the low-and middle-income countries [1].The latest data from WHO and UNICEF showed that two billion people had lack access in managing safe drinking water at home and 800 million of them did not have basic drinking water service [2].Indonesia itself was ranked 9 th out of 10 countries with population that has safe access to 1263 (2023) 012036 IOP Publishing doi:10.1088/1755-1315/1263/1/012036 2 drinking water in ASEAN countries [3].Water insecurity leads people to resort unsafe water sources for their water-related domestic needs [4,5].
Water source is a key factor for community's health [6].The availability of an adequate and uninterrupted clean water supply to the public are directly related to the protection of public health [7].While, the water source contamination is still an issues mostly in low-and middle-income countries, the consumption of contaminated drinking water remains one of the most significant causes of ill in the worldwide [8].The contaminant in the water has routes that can be described and predicted through a human exposure model (HEM).
HEM is now a well developing field of study.It is an important model to predicted the amount of the contaminant that carried by water or other environmental pathways to the community.Due to the competence output of this modelling, proper inputs play a crucial role.In this study, the input of exposure model was identified from the water usage pattern of the communities.It is an important aspect not only to run the exposure model, but also to assess the risk of contaminated water for human health in wider utilization, later on.The problem is, there is limited information about the water usage pattern in a certain area, especially in low-and middle-income countries like Indonesia.
To fill this gap, a water demand survey was done in Kampong Ciwalengke.Kampong Ciwalengke is one of the slums area in the Upper Citarum River Basin (UCRB).The population density in this area is not followed with adequate sanitary facilities and clean water service system.Moreover, Ciwalengke was not included yet in the regency government's clean water master plan.The lack of access to a clean water service system had caused the residents in Kampong Ciwalengke to use unprotected water sources, including Citarum River water [4].The river water was flowed into the village trough pipes and stored in the tub or shallow well.They used this water for their daily activities, even the water did not fulfill the criteria of water source quality based on Indonesia Regulation [9].
The water usage pattern is often complex because it depends on the water source reliability, accessibility/ location and communities economic condition [10].In the implementation, collecting community's water usage data was very challenging as the water demands were varied between individuals.The main objectives of this article are to analyze the resulting variables from the water demand survey and determine whether there are significant differences between groups (i.e., male and female groups, 18-65 years old and ≥ 65 years old groups) or not.This study also aims to distinguish which exposure routes and water-related activities contribute the most to the total uptake of contaminant exposure from the Citarum River.This article provides an overview of how and how much a contaminant could expose people through their daily water use.

Water demand survey
The water demand survey was conducted in Kampong Ciwalengke, a peri-urban area in Upper Citarum River Basin.Kampong Ciwalengke is located in Majalaya District, Bandung Regency (Figure 1).About 60% of residents in Kampong Ciwalengke were used unimproved water sources such as refill water kiosk, unprotected dug well, unprotected dug well with mixed water, and also river water [4].In total, 217 respondents from different households were interviewed.The questionnaire consisted of 73 questions: 14 about personal data and 59 about water usage pattern based on 10 water-related activities of the population study.An ethical clearence agreement with approval number of Etik.RSD/029/I/2019 was issued by the Ethical Committee of Dustira Hospital (Cimahi, West Java) in order to protect the secrecy and rights of the survey respondents.The details of water demand survey design were described in Utami et al [4].

Deterministic HEM equation
This study modelled the contaminant uptake from river to human using deterministic model.Based on Utami et al. [4], the contaminant exposure uptake from Upper Citarum River were distinguished into two routes, i.e. oral and dermal.The oral uptake was calculated by NORMTOX equation from Ragas & Huijbregts [11]: where,   is the daily oral contaminant uptake (ng/kg/day),   is the contaminant concentration in surface water (ng/L), the used contaminant in this study was carbofuran with concentration of 2.6.E+01 ng/L [12].  is the contaminant absorption factor in the gastrointestinal tract (dimensionless), the   value is 1 for carbofuran [13]. is body weight (kg; [4]) and   is the daily water intake rate from water-related activity (L/day).For water-related activities that causing direct ingestion (e.g., drinking water and water for cooking),   was calculated using equation (2): with  is the ingestion volume (L/event; Table 3), and  is the water-related activities frequency per day (event/day; Figure 3).Meanwhile, for water-related activities that causing indirect water ingestion (e.g., swimming, bathing, and brushing teeth),   was calculated using equation (3) below: where,  as ingestion rate of each water-related activity (mL/min; [4]) and  is the exposure duration (min/event; Table 2).The exposure via dermal uptake was calculated by NORMTOX equation that was modified by water dermal contact dose equation [11,14], which can be seen in the equation (4) below: with   as the daily dermal contaminant uptake (ng/kg/day),   is the surface area of the body (cm 2 ; calculated by Mosteller's equation [15]),   is the fraction of body area which contact with the water contaminant (dimensionless; [4]).  is the dermal contaminant absorption rate (cm/hour), the   value for carbofuran is 3.11.E-03 cm/hour, derived from Log   equation from USEPA [16]. is the exposure duration (hours/event; Table 2), and  is the conversion factor (0.001 L/cm 3 ).The total uptake (  ; ng/kg/day) was calculated by summing up the oral and dermal uptake (equation 5).

Statistical and proportion analysis
The data in this study were divided into gender-based and age-based groups.The age groups consisted of two categories, i.e. 18-65 years old and ≥ 65 years old.Statistical analysis of carbofuran uptake was performed using JASP ver.0.17.1.Normality was confirmed by the Saphiro Wilk test.Since, the normality test result did not comply with the normality assumption, so non-parametric tests of Kruskal-Wallis and Dunn's Post Hoc Comparisons were performed to compare the differences of total uptake among all water-related activities.A t-test was performed to compare dermal uptake with oral uptake and the total uptake between groups.The criterion of significance was set at p < 0.05, indicated significant different among the data, and p < 0.001 as highly significant different.Lastly, two proportional analysis were performed to identify the most contribute activity and exposure route for carbofuran total uptake.

Contaminant exposure routes from Citarum River water
This study was started by designing the questionnaire as an efficient tools to gain the real information about community's water usage as accurate as possible.Before designing the questionnaire, we observed the population behaviors regarding their daily activities involving the use of water in the study area.Observation of the respondents' daily activities before the questionnaire survey was needed to ensure that respondents' answers were close enough to their real condition.Secondly, we did literature studies about the river water used by the population in river basin, especially in Citarum River Basin and what kind of common water-related activities were done by the peri-urban community.Those two actions resulting exposure routes diagram in Figure 2 as a baseline of the questionnaire survey design.
Figure 2 shows that contaminants are released into the aquatic ecosystem through various anthropogenic activities, such as agriculture, industries, aquaculture, husbandry, and domestic hygiene activities.Furthermore, these contaminants may also expose humans through various pathways, such as drinking water consumption, consumption of water-used products (e.g.agricultural, aquaculture, and husbandry products), domestic water-related activities (e.g.religious activities, cooking, and hygiene activities), water sport, and mining.Exposure pathways of contaminants in Citarum River lead to contaminants uptake through the skin (dermal) and gastrointestinal tract (oral) [4].
Typically, clean water is used for drinking, hygienic activities, cooking and food preparation, and other water-related indoor and outdoor activities that happen around the house and homestead such as poultry rearing, washing car and gardening [17].Apparently, from the water demand survey, 60% of respondents used unsafe water sources for these personal and domestic needs [4].This might be due to a lack of awareness and knowledge about the water safety [18].

HEM variables from water demand survey
In this study, we created a database containing specific information regarding the water usage pattern of peri-urban community in the study area, consisting of frequency, volume, and duration of 10 waterrelated activities.Based on the provided data on Table 1, it seemed that the participation rates were quite high across all activities, as indicated by the 100% values for drinking water, teeth brushing, bathing, ablution, and hand washing.These results suggested that those activities were universally practiced by all respondents.Whereas, the percentages of male and female users were generally consistent across most activities, with the percentages of male group were slightly lower than female group.From the percentages, indicated that both genders were equally involved in water-related activities.However, it could not be concluded as simply as that, since the number of respondents between gender groups were quite different (i.e.175 females and 42 males).This difference was also due to the survey's timing.The survey was executed during the daytime (08:00 -18:00) when mostly male residents were at work.The percentages of respondents who did the water-related activities were presented on Table 1.However, for activities such as laundry, cooking, and food washing, the participation rates dropped slightly for respondents aged ≥ 65, indicating a potential difference in preferences or needs between the age groups.The water-related activity with the lowest participation rate was swimming, with only 0.92% of respondents engaging in it.This suggested that swimming was not a commonly practiced waterrelated activity among the surveyed individuals.While the majority of respondents engaged in laundry and dish washing, there were slightly lower participation rates (99.54% and 98.16%, respectively) compared to other activities.This could be attributed to variations in personal circumstances, such as the availability of washing facilities or the use of alternative methods [10,19].This result highlighted the consistent participation in essential water-related activities such as drinking water, hygiene practices (teeth brushing, bathing, ablution, and hand washing), as well as household tasks (laundry, dish washing, cooking, and food washing).The lower participation rates in activity like swimming may be influenced by many factors such as personal behavior, availability of resources, or specific demographic factors [10,19,20].
To predict the contaminant exposure uptake from river water to human, we need basic data of volume, duration and frequency of the water usage for each person.The first resulted variable from the water demand survey was frequency of each water-related activities.Figure 3 showed that only swimming and ablution activities that had same frequency among all respondents, while other frequencies of water-related activity were varied ranging from 1-2 times per week to > 3 times per week.The frequencies of water-related activities would influence the water intake rate for each individual.However, in term of HEM, only water consumption from oral route activities could be directly used as a variable for daily intake rate (  ; L/day) to calculate the contaminant uptake.Those water-related activities were drinking water and cooking.While, for other activities, we could determine the daily intake rate by using the duration variable.Higher uptake of a water-related activity might be due to longer duration of exposure from contaminated water.Table 2 showed the average duration of each water-related activity. 1 -1 1 -1 1 -1 1 -1 1 -1 Hands washing (N=217) 0.5 -0.5 0.5 -0.5 0.5 -0.5 0.5 -0.5 0.5 -0.5 Food washing (N=207) 2 -2 2 -2 2 -2 2 -2 2 -2 Swimming (N=2) 60 -120 -60 -60 120 -120 60 -120 The longest duration of water usage was in laundry activity with average duration of 25.4 (±14.36)min/person/event, with a range of 10 to 120 minutes.This suggested that laundry task require a substantial time investment.Dish washing had an average duration of 8.84 min/person/event, with a range of 2 to 20 minutes.This implied that dish washing was generally completed within a moderate timeframe.While the shortest duration of water usage was in hands washing activity with average duration of 0.5 min/person/event.This indicated that this activity was typically quick and efficient.Activities that did not have duration variable were drinking and cooking.The data for swimming was based on a very small sample size of only 2 respondents.Therefore, the average duration of 90.00 min/person/event should be interpreted with caution.It was worth noting that the duration range for swimming was between 60 and 120 minutes, indicating a moderate to long duration for this activity.Duration per one occurrence or per event of ablution, food washing, and hand washing activities were assumed as uniform for every person.The average durations and duration ranges for the various activities did not show significant variations across age groups (18-65 and ≥ 65) or between genders.This suggests that the average time spent on these activities is relatively consistent regardless of age or gender.
The results suggested that respondents spent varying amounts of time on different water-related activities.While some activities such as teeth brushing and ablution were completed relatively quickly, others like bathing and laundry require more substantial time investments.The average data did not indicate any significant differences in activity duration based on age groups or gender.Another variable resulting from the questionnaire survey is water volume which is presented on Table 3.There was significant variation in average water volume across different water-related activities.
Activities like bathing and laundry generally require larger volumes of water compared to activities like drinking water or cooking.The highest average water volume per activity event was during bathing with 28.1 (±15.90)L/person/event followed by laundry with an average of 20.29 (±13.82)L/person/event.Meanwhile, activities like drinking water and cooking have relatively low volumes.The lowest volume was in drinking with average volume per activity event was 0.2 (±0.10)L/person/event.The range of water volume per event did not only vary across activities, but also showed a wider range for certain activities.For example, a wide volume range was indicated for bathing activity with 6.3 -137.52 L/person/event, suggesting that there was variability in water consumption patterns among respondents.These informations can be valuable for understanding water usage patterns, providing insights on the community behavior, planning water resource management, and promoting water conservation practices.

Uptake comparation between groups
From the variables of frequency, duration and volume from the water demand survey, the contaminant uptake can be estimated using deterministic HEM equations.Contaminant concentration (  ) that we used in the model was carbofuran (2.6.E+01 ng/L), since this contaminant ranked at the top of the priority list for health risks in the UCRB [12].The average uptake of carbofuran based on exposure routes and groups were showed in Table 4.The carbofuran total uptake based on exposure routes showed that the exposure from oral uptake (1.05.E+00 ng/kg/day) was significantly higher than the dermal uptake (1.14.E-02 ng/kg/day).The amount of exposure from dermal uptake is considered low or negligible compared to the oral uptake.The statistical t-test comparing the two uptake routes also suggested consistent results.As shown in Table 5, the p-value for the comparison of exposure routes was 8.51.E-90 (p < 0.001), indicating a highly significant difference between dermal uptake and oral uptake data.The total uptake of carbofuran for both age categories were 1.05.E+00 ng/kg/day for age category 18-65 years old and 1.28.E+00 ng/kg/day for age category ≥ 65 years old.Higher uptake for ≥ 65 group was probably caused by lower body weight value of elder people.Meanwhile, carbofuran total uptake for gender groups were 1.07.E+00 ng/kg/day for female and 1.04.E+00 ng/kg/day for male.This result indicated that the contaminant uptake was higher for female than male group.This might be happened because of sociocultural norms and roles may influence how females and males interact with water sources and their behaviors concerning water-related domestic activities [21].In the study area, it's common for women to responsible with all domestic water-related activities rather than men.This gender disparity is also a significant concern in WASH fields, especially in low-and middle-income countries [22].In the context of Ciwalengke and Indonesia, it is an undeniable reality that women play a vital role in domestic water-related activities and are, consequently, more susceptible to exposure to contaminants from water sources due to their responsibilities [21,23].Previous studies suggested that the different gender and/ or age could have a significant impact on the contaminant uptake [21,24].The t-test result of total uptake (ng/kg/day) between groups in this study are showed in Table 5.The variance of the total uptake for the age group 18-65 was 0.18, whereas for the age group ≥ 65 it was 0.3922.This indicated that there was a higher variability in the total uptake among individuals in the ≥ 65 age group.The number of observations for the age group 18-65 was 200, while for the age group ≥ 65 it is 17.This signified that there was a larger sample size for the 18-65 age group, providing more reliable data.The two-tailed p-value for the t-test was 0.1518.This represents the probability of observing a t-statistic as extreme as the calculated value in either direction, assuming there was a difference in total uptake between the two age groups.In summary, the t-test suggested that there might be a difference in total uptake between the age group 18-65 and the age group ≥ 65, although the difference was not statistically significant at the conventional significance level (p > 0.05).
For the statistical analysis of carbofuran total uptake between gender, the variance represented the spread or variability of the data.In this case, the total uptake in males had a variance of 0.1101, while in females it is 0.2207.This suggested that there was less variability in the total uptake among males compared to females.The number of sample size (observations) for males was 42, while for females it was 175.It could be said that females group indicated a more robust dataset.The two-tailed p-value was 0.6437.Based on the that result, there was no significant difference observed in the total uptake between males and females (p > 0.05).The calculated t-statistic was not statistically significant, as it did not exceed the critical values at the chosen significance level.
The output of deterministic model also classified into total uptakes from various water-related activities that is shown in Table 6.It could reflect amount uptake of contaminants related to the water demand of each activity.The highest total uptake were from drinking water, water for cooking, and bathing, respectively (Table 6).These three activities were considered higher than other activities.Statistical tests of Kruskal-Wallis and Dunn's Post Hoc Comparisons were performed to determine if there are statistically significant differences of carbofuran total uptake among water-related activities.
The Kruskal-Wallis p-value result was < 0.001, indicating a highly significant difference between all activities.Meanwhile, the Dunn's Post Hoc Comparisons tests resulted 45 pairs of water-related activities comparisons with 33 pairs of activities showed highly significant differences (p < 0.001), 4 pairs showed significant differences (2 pairs with p < 0.01 and 2 pairs with p < 0.05), and 8 pairs showed no significant differences (p > 0.05).Those 8 pairs activities which showed no significant differences between its carbofuran uptakes were, i.e. ablution-swimming, bathing-swimming, cooking-swimming, dish washing-laundry, dish washing-swimming, drinking-swimming, laundry-swimming, and swimming-teeth brushing.From these results, we can conclude that almost all carbofuran uptake from different activities were significantly different between each other.

Variables proportional analysis
The proportional analysis was performed to assess which route and activity that contribute the most to the contaminants total uptake.From the exposure routes proportional analysis (Figure 4a), it could be observed that oral exposure contributed almost 99% of the total uptake.Meanwhile, the proportional analysis of water-related activities' uptake showed that drinking water and cooking are the two activities that contribute the most to the total uptake of carbofuran, with drinking water being the highest (Figure 4b).With a contribution of approximately 77.59%, drinking water was the most significant activity in terms of water uptake.Cooking accounted for about 20.6% of the total uptake.This suggested that water usage in food preparation, such as boiling, steaming, or washing ingredients, played a quiet significant role in the overall contaminant uptake from water usage.Bathing contributed approximately 2.5% to the total uptake.It indicated that bathing, although essential for personal hygiene, accounted for a relatively smaller proportion of carbofuran uptake compared to drinking water and cooking that had bigger proportion than others.The proportional analysis result for total uptake is shown in Figure 4. Other activities had relatively low contributions compared to drinking water and cooking.Activities such as teeth brushing, laundry, dish washing, ablution, hands washing, food washing, and swimming had relatively lower contributions to the total uptake.However, they still played a role in overall water consumption, albeit to a lesser extent.While swimming only contributed approximately 0.605% to the total uptake, it was worth noting as a specific activity with its own water usage.

Conclusion
The water demand survey resulting HEM variables of frequency, duration and volume of water-related activities.Those variables were used in a HEM to calculate the total uptake of carbofuran contamination from the Citarum River.The average total uptake from all water-related activities through oral and dermal routes was 1.06.E+00 ng/kg/day.The statistical analysis of total uptake showed that there was a significant difference in total uptake between oral routes but no significant difference between gender or age groups.The proportional analysis of the uptakes indicated that oral uptake contributed 98.8% to the total uptake.The analysis also revealed that that drinking (77.6%) and cooking (20.6%) were two activities that contributed the most to the total uptake of carbofuran exposure.Other activities had relatively low contributions compared to drinking and cooking.Overall, the analysis emphasized the importance of selecting safe water source for drinking and cooking activities since the contribution from these two activities were highest among all.

Figure 1 .
Figure 1.Study location of Kampong Ciwalengke in the Upper Citarum River Basin, West Java.

Figure 2 .
Figure 2. Contaminant exposure pathways from Citarum River to the community.

Figure 3 .
Figure 3. Frequency of water used among respondents.

Figure 4 .
Figure 4. Proportional analysis results of carbofuran uptakes between exposure routes (a) and waterrelated activities (b; in lognormal).

Table 1 .
The percentage of respondents for each water-related activity based on the questionnaire survey (number of total respondents [N] = 217).The percentage data in Table 1 also included information on different age groups.For activities like drinking water, teeth brushing, bathing, ablution, and hand washing, the participation rates were consistent among respondents aged 18-65 (200 respondents) and those aged ≥ 65 (17 respondents).

Table 2 .
Average duration of water-related activities per event.

Table 3 .
Average water volume of water-related activities per event.

Table 4 .
Average carbofuran uptake based on exposure routes.

Table 5 .
T-test results of carbofuran uptake between groups and oral routes.

Table 6 .
Average carbofuran uptake based on water-related activities.