Influence of Catchment Surface Material on Quality of Harvested Rainwater

Abstract

Rainwater harvesting (RWH) systems offer an opportunity to diversify the water matrix under drought scenario. However, there is not a robust institutional framework for managing this new source of water. The objective of this study was to evaluate the influence of the catchment surface material on the quality of rainwater collected. Three systems were installed in south-central Chile, which collected rainwater from three different surfaces: gravel tile, zinc-polycarbonate sheets, and clay tiles. The RWH system consists of a first flush diverter and then a storage system with chlorination. The rainwater was characterized by its physicochemical and microbiological properties for its possible final use, considering the current regulations for drinking water and agricultural irrigation. The results indicate that the raw rainwater harvested from different surfaces presented a high mean conductivity of 232 ± 146 µS/cm. Meanwhile, fecal coliform values were <1 MPN/100 mL, which indicates good microbiological quality. Regarding the quality for use as drinking water, it was found that of 39 parameters evaluated according to a reference standard, only the pH was above the indicated limit. Meanwhile, the harvested water presents satisfactory quality for irrigation, except for its sodium (35–50% above the standard) and pesticide values (with respect to 0.028 µg Delta-BHC/L at Site 2).

1. Introduction

Climate change is directly influencing water availability around the world. Projections indicate that by 2050, more than 50% of the world population will live in areas under conditions of water stress because of the physical lack of supply and the inability of management systems to meet human and ecological demand [1]. The diversification of water sources and the development of infrastructure to provide water from nonconventional sources is an alternative that must be considered to foster resilience amid global change [2]. Against this backdrop, the implementation of rainwater harvesting systems arises.

Rainwater harvesting (RWH) systems generally consist of ways to harvest and store rainwater and prevent its runoff. RWH involves the collection, storage, treatment, and distribution of rainwater from roofs, rooftops, and impermeable surfaces for on-site use [3]. The uses of collected rainwater include toilet flushing, laundry, garden irrigation, patio cleaning, and car washing, among others. These uses are aimed at reducing the consumption of drinking water from centralized sources [4]. At an application level, Germany and Australia are leaders in the implementation of these systems; the latter had around 1.7 million homes with rainwater tanks as of 2014. In South America, Brazil stands out, with its One Million Cisterns program, and in Central America, the Isla Urbana initiative in Mexico City has allowed the collection of 170 mL of runoff from roofs to alleviate local water scarcity and flooding problems [4]. In Chile, the prolonged drought during the 2010–2014 period led the authorities and technical organizations to consider promoting the use of RWH in rural areas and, in relevant cases, in urban zones [5].

Rainwater harvesting can contribute to covering the drinking water supply in urban and rural areas, but it is very important to have a defined institutional framework that can regulate its management, monitoring, and use [6]. This new resource is an important opportunity to provide water for demanding uses that do not require high quality [7]. These uses can account for 80–90% of water consumption in homes [8]. Rainwater use contributes to water self-sufficiency in cities, decreasing pressure on conventional water supply systems [9].

Agricultural use is another alternative for harvested water. This water is a complement in periods of water scarcity or stress in areas where precipitation is insufficient for crop growth [10]. Rainwater harvesting can also mitigate the impacts of the intense rainfall that can occur under climate change, reducing runoff in impermeable urban areas [11] and/or soil erosion and degradation [12] in rural areas. In the latter case, it contributes to agricultural development and resource conservation [13].

Among the key factors in rainwater harvesting are the building roofs or catchment areas used for rainwater conveyance and storage. The materials of these surfaces have a strong influence on the quality of the collected rainwater [14]. Thus, rainwater quality is determined by the atmospheric conditions of the air where the rain is produced and by the surfaces and infrastructure where water collection takes place.

Precipitation events are increasingly limited and uncertain and depend on the geographic region. It is also important to ascertain the quality of the collected water to generate a robust institutional framework that regulates the safe use of harvested rainwater. The objective of this study is to assess the influence of surface material on harvested rainwater quality. The information gathered in this study is crucial for building an institutional framework to appropriately manage the use of rainwater harvesting.

2. Materials and Methods

2.1. Study Site

Three RWH systems were installed in three different homes, all located in Concepción, Biobío Region, Chile (36°46′22″ S latitude and 73°03′47″ W longitude). The climate is temperate, with an average cumulative annual precipitation of over 616 mm, according to data recorded between 2016 and 2020. In the same period, average precipitation was concentrated between the months of April and October and exceeded 50 mm/month/year [15]. Figure 1 shows the geographic location of the three installed systems, a photograph of each, and the average cumulative annual rainfall in the country between 2016 and 2020.

2.2. Rainfall Quantity

To have a reference regarding feasibility in terms of the quantity of rainwater possible to collect in the Concepción Province, data from the 2016–2020 period were compiled. The data were grouped monthly, with average cumulative monthly precipitation per year in the mentioned period obtained in accordance with data from the Chilean Meteorological Service [15].

2.3. RWH System Design

The three installed systems have similar designs and are distinguished by the catchment surface used to collect rainwater from the roofs of homes. RWH system design incorporates a physical and disinfection treatment unit and a final water quality monitoring unit.

Water catchment occurred via the roofs of homes made of different materials: (a) gravel tile (Site 1), (b) zinc-polycarbonate sheets, and (c) clay tiles (Site 3).

Table 1. Description of rainwater harvesting systems and treatment of harvested water.

Section	Rainwater Harvesting System
	Site 1	Site 2	Site 3
Roof material	Gravel tile	Zinc-polycarbonate sheets	Clay tiles
Housing area (m2)	50	25	94
Connected rainwater downspouts (material)	3 (PVC)	2 (PVC)	1 (tin and PVC)
Leaf filter (stainless steel)	1	2	1
First flush diverter (L)	300	18	300
Storage tank (L)	2700	3400	2700
Disinfection	Calcium hypochlorite	Calcium hypochlorite	Calcium hypochlorite
Filter pore diameter (µm)	130 and 5	130 and 5	130 and 5

details the characteristics of the catchment surfaces (roofs and pipes), the disinfectant used, and the pore size of the filters installed prior to the treated harvested rainwater use point.

2.4. RWH System Description

The rainwater was captured by the different roof surfaces and conveyed through gutters to a stainless-steel mesh filter to retain leaves and larger solids. Then, the first millimeters of rain were conveyed to the first flush diverter to prevent particles, sediment, coloring, organic matter, and even pathogens that had accumulated on the roof, as well as pesticides and organic compounds entrained by the rain from the air, from reaching the water storage tank [14,16]. Once the rainwater reached the storage tank, it was disinfected by dosing with calcium hypochlorite (65% strength) in accordance with local regulations and the intended use of the harvested water. Finally, the water was extracted with a 0.5-HP pump to then pass through two filters with different pore sizes (130 µm and 5 µm) and harvested at the outlet tap for its end use. Figure 2 shows a scheme that represents the three installed systems.

2.5. Water Harvest Sampling

The rainwater sampling was conducted in the winter of the region (May–August 2022). Samples were taken from two sampling points: the first flush diverter to characterize the raw water entering the system and from the outlet tap to characterize the harvested water in each of the studied systems. The samples were transported in coolers with ice packs and subsequently stored at 4 °C in darkness until their assessment.

2.6. Analytical Methods

The quality of the harvested water samples was determined according to Chilean drinking water standard NCh 409/1.Of2005 [17] and irrigation water quality standard NCh 1333.Of78 [18]. The physicochemical and microbiological analyses used to assess raw and harvested water quality were performed following Standards Methods [19]. In addition to determining rainwater quality, the results were compared with the Australian drinking water standard [20] and international guidelines from the USEPA and WHO [21,22].

3. Results and Discussion

3.1. Rainfall Quantity in Concepción, Chile: 2016–2020

According to Ako et al. [23], to determine if rainwater collection is an appropriate technology, the first task is to estimate the potential supply of rain to ensure that household needs are met. Figure 3 shows the average cumulative monthly precipitation by year in the 2016–2020 period in Concepción (Chile).

The cumulative annual precipitation in the study area is in the range of 616–930 mm, with an average between 2016–2020 of 770 ± 116 mm. This value is 90–93% greater than the average precipitation per year. This is because there are months in which precipitation reaches 345 mm (June 2020), meaning 80.5% greater than the annual average, but there are others in which this is 100% greater than monthly precipitation (February 2016 and 2019, November 2020, and December 2019).

This variation in the distribution of precipitation throughout the year reinforces the importance of being able to store rainwater for drought periods. Considering rainwater harvesting and climate change, it is expected that in temperate zones, the increase in water security will reach between 77–83% reliability, which would increase in homes with a larger catchment area and storage volume [24]. Domènech and Saurí [25] show that the implementation and use of home RWH systems in the city of Sant Cugat del Vallès, Spain, accounted for a 16% reduction in drinking water demand. Vieira and Ghisi [26] determined that in Florianópolis, Brazil, RWH systems were able to meet an average of 43% of the drinking water demand during the study period. Meanwhile, Alim et al. (2020) [27] assessed the meeting of the need for water for potable and non-potable uses of a small-scale system in rural communities in Australia, obtaining 90–97% reliability. Another important factor is storage tank volume, as a greater volume will increase system reliability in periods with greater precipitation. However, it will also increase investment costs and the probability that there will be a small volume of water stored during dry periods [28]. Finding the ideal storage volume for the local conditions of the study area is vital for ensuring RWH system sustainability. So, rainwater harvesting can contribute to the diversification of water sources and put rainwater to sustainable use, providing water security [29]. Under conditions of climate change, the number of intense rainfall events will increase, but there will be a greater number of dry periods, which can affect public water supply systems [30], and deficits could be partially covered by such technologies. Moreover, rainwater harvesting in countries such as China, Brazil, and Argentina supplies activities such as livestock raising, small-scale irrigation, and even water table recovery. While in Thailand, 40% of the population depends on rainwater harvesting as a water supply mechanism [31].

3.2. Raw Rainwater Quality from Different Catchment Surfaces

Below, the quality of the raw water extracted directly from the rainwater diverters in the RWH systems at the three study sites, based on in situ and physicochemical parameters relevant to this study, is detailed. Table S1 of the Supplementary Materials includes quality values of the raw rainwater collected from roofs made of different materials that have been reported in the literature and those from this study.

3.2.1. In Situ Parameters

Electrical conductivity is in the range of 103–390 µS/cm, with an average of 232 ± 146 µS/cm among the three study sites. This value is 61–68% greater than those reported in previous studies [32,33]. The average conductivity obtained in this study is consistent with Martínez-Castrejón et al. [34], who obtained 221 µS/cm in a water sample from a first flush diverter in a system whose catchment material was galvanized metal. This parameter is very correlated with the dry period prior to the rain catchment period [32], and it could be correlated in areas near marine coasts due to saline aerosol transport [33]. Meanwhile, conductivity values can increase due to dissolved organic solids composed of ions such as chloride, sodium, nitrates, and phosphates deposited on roofs and leachate from construction materials on the catchment surface [35,36]. Adedeji et al. [35] show that there is an average difference of 38% in conductivity between a new (88.7–100 µS/cm) and rusted galvanized iron roof (145–160 µS/cm), which highlights the importance of the state of the roof in collected rainwater quality. In addition, the conductivity of the rainwater increased by 71% after being harvested from an asbestos roof, demonstrating the effect of material type on raw rainwater quality.

Meanwhile, the pH of the first flush of water obtained from the diverter that received water from the roofs of the three studied sites was between 6.12–6.74, with an average of 6.38 ± 0.32. The average pH is very close to that reported by Ako et al. [23], who studied ten samples of rainwater harvested from roofs made of corrugated iron sheets in Yaoundé, the capital of Cameroon, obtaining a value of 6.25 ± 0.45. Tengan and Akoto [37] found a significant difference between the pH of water collected directly (pH: 4.92 ± 0.29) and the pH values of water harvested via roofs made of aluminum, asbestos, and galvanized, which reached 5.33 ± 0.09. The lower pH of the raw rainwater favors the solubility of metals in water, increasing their availability due to erosion caused by the interaction of rainwater with roof surfaces, increasing dissolved solids content, and generating a replacement of hydrogen ions with metal ions [37]. Furthermore, pH has a great influence on corrosion rates, and the Langelier saturation index (LSI) is one of the most used. Maeng et al. [38] determined that there was a decrease in pH during the rainy season, as well as in the LSI, which corresponds to a more corrosive environment.

3.2.2. Organic Matter and Nutrients

In this study, the samples from the rainwater diverters, regarding nutrients, presented values below 1 mg/L of total nitrogen (TN) (from the three study sites). For organic matter, the samples presented values of 0.91–1.21 mg/L of total organic carbon (TOC), and below 2.5–10.1 mg/L of chemical oxygen demand (COD). The rainwater from the roof of Site 2 (zinc and polycarbonate) presented the highest values, with 1.21 mg/L of TOC and 10.1 mg/L of COD. Similarly, Friedler et al. [33] found that materials such as concrete and tile have depressions and cracks that substances can adhere to more easily, making it difficult for them to be entrained by rainwater. However, smooth surfaces tend to increase parameters such as TN, TOC, and COD. Meanwhile, Zhang et al. [32] reported finding that the longer the dry period prior to rain, the more the total nitrogen concentration, TOC, and COD, in water collected from concrete, asphalt, and tile roofs increased, probably due to the greater accumulation of organic waste on catchment surfaces. Another nutrient that was assessed was N-NO₃, with values of 0.05, 0.13, and 1.29 mg N-NO₃/L (equivalent to 0.22, 0.58, and 5.7 mg NO₃/L, respectively) for water from clay tile, zinc and polycarbonate, and gravel tile roofs, respectively. The average concentration was 2.18 ± 3.6 mg NO₃/L. Likewise, Martínez-Castrejón et al. [34] observed a value of 2.1 mg NO₃/L. In another study, led by Lee et al. [39], NO₃ concentrations equivalent to 1.89, 2.55, and 2.80 mg/L in water from roofs made of clay tiles, concrete tiles, and galvanized steel sheets, respectively, were reported. The authors explain that these values are expected due to the growth of mosses and lichens, whose microbiological processes increase nitrate and sulfate concentrations.

3.2.3. Suspended Solids and Turbidity

Regarding suspended solids and the turbidity of the raw rainwater, Figure 4 shows the concentrations of total suspended solids (TSS) and volatile suspended solids (VSS) and turbidity of the rainwater from the different catchment surfaces in this study.

The SST concentration varied between 10.5 mg/L and 23.4 mg/L, with an average of 15.3 ± 7.03 mg/L in the water from the diverters of the three studied sites. Gikas and Tsihrintzis [7] characterized water collected in diverter tanks installed in different rural, suburban, and urban zones of Greece, observing total suspended solids concentrations between 9.5 mg/L and 39.5 mg/L, with the highest concentration in a rural zone near an industrial area. However, they note that the concentrations in the first flush diverters were 2.4–15.2 times greater than in the storage tanks after the diverter, which they attribute mainly to accumulated solids on the roofs during the period before the rain. VSS presented values between 5.35 mg/L and 19.9 mg/L, and at Sites 1 and 3, they accounted for 51% and 50% of TSS, respectively. However, at Site 2 this figure reached 85%, pointing to the organic nature of the solid deposits on the catchment surfaces entrained by the first rainwater. According to average solids data reported by Friedler et al. [33], this proportion reached 69%, considering that the ranges reported for total suspended and volatile solids were 0–279 mg/L and 0–244 mg/L, respectively, determined in a total of 153 samples.

Turbidity was between 0.24 and 1.07 nephelometric turbidity units (NTU). These results are similar to the findings of Fitobór and Quant [40], whose study in a tourist area surrounded by farmland and forests in Poland concluded that in all their samples from the first phase of rain, this parameter did not exceed 4 NTU. The authors stated a high suspended solids concentration is closely related to the level of harvested rainwater contamination, as the suspension is an excellent contaminant carrier. This is consistent with the higher total suspended solids and volatile suspended solids concentrations, turbidity, and total coliform concentrations observed in the first batch of rainwater at Site 2, equal to 23.4 mg/L, 19.9 mg/L, 1.07 NTU, and 2.7 most probable number (MPN)/100 mL, respectively. Lye [41] even states that there is a positive correlation between suspended solids and turbidity with a greater concentration of organic substances (including oil-based products), nutrients, microorganisms, and even heavy metals. Regarding total (except at Site 2) and fecal coliform content, all the samples presented values < 1 MPN/100 mL, indicative of a low microbiological concentration in the first batches of rainwater at the studied sites.

3.2.4. Ions

Figure 5 shows the ion concentrations in the raw rainwater from the different catchment surfaces in this study.

Chloride concentrations varied between 0.28 mg/L and 0.41 mg/L, with an average of 0.36 ± 0.07 mg/L, within the range reported by Nnaji and Nnam [42]—between 0 mg/L and4 mg/L—in a study focused on characterizing rainwater from 25 systems installed in Nigeria. Regarding fluoride, the concentrations were between 0.21 mg/L and 0.35 mg/L, which is consistent with the findings of Zhang et al. [32]. However, these authors highlighted that of all the samplings carried out, 2.8%, 6.4%, and 10.6% of the time, the fluoride concentration exceeded 1 mg/L in water from roofs made of concrete, ceramic, and asphalt. Meanwhile, the sodium and potassium values observed in this study were 1.55–2.06 mg/L and 0.54–0.95 mg/L, respectively. Lee et al. [14] found similar ranges in water harvested from a PCV tank and galvanized and aluminum gutters roofs on buildings in the northwest of South Korea. Finally, the average sulfate concentration in this study was 0.99 ± 0.11 mg/L, a value below but near the 1.57 mg/L and lower than the 2.87–3.64 mg/L reported by Dobrowsky et al. [43] and Lee et al. [39], respectively.

3.3. Harvested Rainwater Quality by Use

A well-designed rainwater harvesting system equipped with technology that allows water quantity and quality monitoring can ensure good-quality harvested water for drinking [44]. In addition, Richards et al. [45] concluded that savings equivalent to 25% of water used for non-potable purposes are possible with these systems, meaning that they could be a way to help address health and sanitation problems in rural zones that are vulnerable to water scarcity. In rural zones in developing countries, a safe drinking water supply is a challenge; therefore, appropriate harvesting and treatment of rainwater could be a method to help achieve equal access to safe, affordable services, as suggested by Sustainable Development Goal 6 [46]. Lee et al. [47] suggested that due to the good quality of rainwater, guidelines should be implemented with specific quality parameters. Moreover, long-term scientific research has established the baseline of potential contamination, according to the clime, catchment surface materials, and RWH system, and has addressed effective system treatment and maintenance aspects. Therefore, below, the quality of the rainwater harvested at each studied site is detailed, and based on limit values indicated in Chile and internationally, possible uses are determined.

3.3.1. Quality of Drinking Water

An important parameter that needs to be determined when assessing drinking water quality is the residual chlorine concentration. According to NCh 409/1.Of 2005 [17], the residual chlorine concentration must be between 0.2 mg/L and 2 mg/L, for which in this study, a dose of calcium hypochlorite (65% strength) of between 3.6 g and 4.3 g was added to the tanks depending on the study site. This dosage had a direct effect on the chloride concentrations of the treated water after filtration (130 µm and 5 µm), with an average concentration of 5.74 ± 0.86 mg/L, equivalent to 96% greater than the average of the water in the three diverters. The chloride concentration in harvested rainwater was 16 times greater than that of the raw rainwater; however, it was below the 400 mg/L and 250 mg/L established as maximum limits by Chilean and Australian drinking water standards, respectively [17,20]. Richards et al. [45] reported a chloride concentration 3.5 times greater than that in this study; considering that these authors added sodium hypochlorite (4% strength), they determined the chlorine dose according to the sum of chlorine demand (mg/L) and residual chlorine (mg/L). However, in this study, an absence of E. coli and total coliforms below MPN/100 mL was observed in all the treated water samples. Only the raw water of Site 2 presented a total coliform concentration above 1 MPN/100 mL (2.7 MPN/100 mL) in the diverter. This indicates good microbiological quality of the water harvested from the three study sites, given that the harvested water was obtained 8 days after calcium hypochlorite addition. Meanwhile, nitrate concentrations in the treated water varied between 0.81 mg/L and 1.22 mg/L. Bui et al. [44] (2021), who applied serial filtration (5 µm and nanofilter) and UV for disinfection, found a final concentration that varied between 0.2 mg/L and 0.3 mg/L. These values are below the national and international limits for the maximum nitrate concentration in drinking water cited in this study (50 mg/L).

On the other hand, Table S2 (Supplementary Materials) shows a characterization of the 39 parameters; 97% of them agree with the values indicated by the Chilean drinking water standard. Only the pH was outside the acceptable range (6.5–8.5) because the pH range at the three study sites was between 5.9 and 6.35. Parameters such as benzene and cadmium showed values less than 5 µg/L and less than 0.01 mg/L, respectively. The Australian standard indicates maximum limits of 1 µg/L and 0.002 mg/L, respectively; as the limits of detection of these two elements are above the values indicated by the Australian standard, the values reported in this study make it difficult to determine their exact concentrations. The same occurs when referring to WHO [22] guidelines, which indicate that the maximum cadmium value in water for human consumption is 0.003 mg/L. Almost all the parameters meet the requirements established for drinking water quality at the three study sites, as reported by Richards et al. [45]. They concluded that a RWH system equipped with low-cost methods of disinfection with sodium hypochlorite produces good-quality water that meets drinking water standards in India. Therefore, it is a technology that contributes to reducing water stress, especially in vulnerable areas. Bui et al. [44] even stated a treatment based on a 5-µm filter, a nanofilter, and UV disinfection (29 W) avoided the addition of chemical products such as sodium or calcium hypochlorite, and drinking-quality water could be obtained. This approach can make rainwater harvesting technology sustainable because it is free to use and will not generate social conflicts.

3.3.2. Water Quality for Irrigation and Recreation

Table S3 (Supplementary Materials) shows the concentrations of 30 physicochemical parameters, including fecal coliforms. Moreover, 97% of these parameters are within the Chilean irrigation water standard, with the exception being the percentage of sodium, which exceeded it between 1.2 and 1.4 times (Site 1).

Of the determined parameters, it bears mentioning that zinc concentrations were between 0.45 mg/L and 1.18 mg/L, a range that does not exceed the limit indicated by the national standard for irrigation water. Serial filtration took place prior to the sample collection in this study, as zinc concentrations between 0.01 mg/L and 12.36 mg/L in raw rainwater harvested from different roofs have been reported in the literature [48,49]. The presence of metals such as zinc and copper is highly associated with the corrosion of construction materials, mainly roofs [50]. The characterization also included organo-chlorine pesticide (total: 19) and pesticide parameters (total: 12), but they were not reported because the national standard for irrigation water indicates that the competent authority must rule. However, all the values obtained in water harvested were under the limit of experimental detection, except for Delta-BHC (Site 1), for which a concentration of 0.028 µg/L was observed; this compound presents long-range atmospheric transport, is difficult to degrade, and has the capacity for bioaccumulation [51]. Regarding the parameters indicated as requirements by the Australian standard for water use for recreational purposes (total: 21), the water harvested at the three sites met 100% [52].

Meanwhile, the guidelines on water reuse published by the USEPA [21] indicate different degrees of restriction according to the values of the determined physicochemical parameters. Regarding conductivity, they indicate no restrictions on the use of water with values below 700 µS/cm, which coincides with the results of this study. However, sodium adsorption ratio (SAR) values between 0 and 3 and conductivity lower than 200 µS/cm would mean a severe degree of restriction because, if the water has lower SAR values (i.e., lower salinity), the infiltration velocity decreases. This latter aspect should be carefully assessed in future studies, as the percent sodium in the water harvested from the three study sites (40.3–49.5%) exceeded the limit established by the Chilean standard for water use in irrigation (35%). Regarding the toxicity of specific ions, the United States guidelines indicate that, with respect to sodium, SAR values below 3 in water mean that it can be used with no restrictions without harmful effects, especially on sensitive crops. However, agricultural crops, like lettuce, can be grown considering the addition of arbuscular mycorrhizal fungi strains under salinity conditions [53,54,55]. The same is the case for pH, boron, and nitrate nitrogen, as all the water samples presented values between 6.5 and 8.4, below 0.7 mg/L, and below 5 mg N-NO₃/L, respectively, which is classified as having no degree of restriction on use on crops susceptible to miscellaneous effects caused by these elements. In terms of metals and metalloids indicated by the United States guidelines, the values of which are 100% consistent with those indicated by the Chilean standard, this study meets in both cases [21].

4. Conclusions

The quality of the first rainwater collected from the surfaces of gravel tile, zinc-polycarbonate sheets, and clay tiles presented conductivity values with a high standard deviation—232 ± 146 µS/cm—and that were above the range reported in the literature (75–91 µS/cm). However, all the samples presented fecal coliform values below 1 MPN/100 mL, indicating that the rainwater collected from the different studied roofs had good microbiological quality.

Regarding the quality of the water collected for drinking water use, of the 39 parameters evaluated according to a reference standard for this study, 97% were below the indicated limit values, with the exception being pH. Assessment of the quality of harvested water treated for use in agricultural irrigation involved the determination of 30 parameters indicated by the standard used as a reference in this study, of which 97% complied, with the exception being the percent sodium value observed in the three samples (values above 35% and between 40% and 50%). Regarding organochlorine parameters (total: 19 evaluated parameters) and pesticides (total: 12 evaluated parameters), all the values presented concentrations below the limit of experimental detection, except for the sample from Site 2, with respect to Delta-BHC, with 0.028 µg/L.

This study demonstrated the potential of harvesting rainwater at homes in terms of quality and quantity in south-central Chile to diversify the water matrix for non-drinking uses (irrigation and recreation). It is also recommended, for future studies, to analyze whether the rainwater collected by the RWH systems is suitable for other uses, for example, cultivation of fish and crops in the aquaponics.