Quality of rainwater harvesting in urban systems: case study in Colombia

Currently, the phenomena of urban growth and increasing population concentrations in cities have led to addressing one of the principal challenges of modern societies: sustainable water supplies. This resource is increasingly recognised as essential and limited because the available quantity and quality of water have been diminished either by natural processes or by environmental pollution and degradation. In addition, there is a high dependence on the water supply network in centralised systems, which presents significant environmental impact implications for the continued development of infrastructure and the treatment of large volumes of water for urban domestic consumption, regardless of whether the land development and resource sustainability of water are highly conditioned to the growth and expansion of urban networks that are as vital as tap water services.

In this sense, the use of non-conventional water sources such as rainwater in urban systems represents an alternative that should be seriously considered to supply water for certain uses (non-potable) and to serve as a source of water supply to various urban areas, thereby increasing the life cycle of the resource and generating an efficient strategy energetically, economically and environmentally.

Currently, there is little research on the urban metabolism and even fewer studies of problems that affect water resources from the perspective of transition to sustainable development. The problem is serious because the population in urban areas is increasing, and planners and managers often face greater difficulties in water supply for urban environments because of either the need for developing new infrastructure or a shortage or irregularity in the water supply, which are compounded by the general context of climate change.

Water recourses from unconventional origins such as rainwater are not irrelevant, and their optimal use could solve some of the problems related to water in urban areas. Currently, rainwater in urban areas is primarily used for discharges of toilets and watering gardens (Hatt et al., 2004).

As a result of the lack of research, the availability of scientific data of good quality associated with rainwater in cities is emerging very slowly due to insufficient physico-chemical and biological characterisation, quantification and environmental assessment, as well as economic and social assessment. Sub-catchment areas (e.g., roofs, streets, parking lots) can capture rainwater of good quality and in various amounts, which allows for its beneficial use.

Sub-catchment areas such as roofs are prime candidates for collecting rainwater because runoff from roofs is often considered clean (Forster 1999) or at least has relatively good quality standards compared with rainwater in surface intakes (Göbel et al. 2007). Despite these claims, there is still some disagreement about the quality of roof runoff. According to Uba & Aghogho (2000), Gromaire et al. (2001), Simmons et al. (2001), Chang et al. (2004), Adeniyi & Olabanji (2005), Melidis et al. (2007) the assessment of quality of roof runoff indicates that it is between good and acceptable for heavily polluted areas.

To Farreny et al. (2011), the quality of rainwater on roofs depends on the type of roof and the environmental conditions (both the local climate and air pollution). Most research on the quality of rainwater on roofs has been conducted in Asia (Appan 2000; Kim et al. 2005a, 2005b), in Central, Eastern and Northern Europe (Forster 1996, 1999; Zobrist et al. 2000; Gromaire et al. 2001; Albrechtsen 2002; Moilleron et al. 2002; Polkowska et al. 2002; Ward et al. 2010), in Southern Europe, specifically Spain (Farreny et al. 2011; Morales-Pinzón et al. 2012a), in the United States (Van Metre & Mahler 2003; Chang et al. 2004) and Oceania (Simmons et al. 2001; Evans et al. 2006; Magyar et al. 2007; Kus et al. 2010). In Latin America, specifically Colombia, some rainwater harvesting research has been performed and reported, such as the studies by Castañeda (2000); Lara Borrero et al., 2007; Torres et al., 2011; and Morales-Pinzón et al. (2012b).

Currently, software called Plugrisost (Gabarrell, et al. 2014; Morales-Pinzón et al. 2014a; Morales-Pinzón et al. 2015) has been developed that defines the potential use of unconventional water sources depending on the quality. Rainwater can be utilised for non-potable uses in washing machines, toilets, irrigation and other applications. Rainwater is also considered a component in the analysis of sustainable housing, specifically in building homes that take advantage of local materials. Morales-Pinzón et al. (2014b) suggest that the use of unconventional alternative sources of water (rainwater and grey water) for domestic applications represents an opportunity to address the shortage of social housing.

In this study, we evaluated the quality of rainwater and its relation to selected variables: the roofing material, deposit material, piping material and amount of precipitation.

We selected six buildings located in the Pereira–Dosquebradas conurbation (Colombia) in which rainwater harvesting systems had been installed to capture and use rainwater. The systems studied are located in the Environmental Sciences Faculty and the Botanical Garden of the Technological University of Pereira, a dry cleaning service company and two residential dwellings.

The climate in the study area is characterised as mild, with a temperature between 18 and 22 °C, and an average rainfall of 2,750 mm/year, which favour the use of rainwater.

We selected different roof types such as cement material (A), which is used in much of the region, polycarbonate (P) materials, which are used for their brightness, and common Zinc (Z) material, which is used for its low cost (Table 1; Figure 1).

The building connections used for rainwater harvesting are derived from the installation of receiving channels for rainwater that falls from roofs and downpipe rainwater that is stored in tanks. The capture system design shown in Figure 2(a) is used in five of the six buildings studied.

However, in the case of D, the rainwater harvesting system collects water from building downspouts and stores the water in a concrete tank buried in the ground that has a storage capacity of 100 m³. Steel pumps transfer the collected rainwater to toilets (Figure 2(b)).

We identified harvesting system storage tanks constructed with different materials: PVC, polyethylene, asbestos, cement and concrete. Based on the monitoring process, it could be observed that almost all of the buildings have covered storage, which minimises the risk of contamination or the consumption of rainwater by insect habitats. The exception was B, in which the storage system was not covered. With respect to storage tanks, most rainwater harvesting systems are constructed with PVC tanks (50% of the systems observed in our study).

The existing rainwater storage capacity in buildings included in our study is presented in Table 1. The results indicate that D is the building that has the most storage capacity, with 100 m³.

Rainwater samples were collected from the storage tanks after a rain event to evaluate the quality of the rainwater. Laboratories in Environmental Chemistry and Food of the University Technological of Pereira analysed the samples. The basic parameters included in the analysis were those recommended by the World Health Organization (WHO), Columbia Decree 475 of 1998 and Colombia Resolution 2115 of 2007 for evaluating the quality of drinking water, such as the concentrations of bicarbonate, total organic carbon, chloride, total hardness, soluble reactive phosphorus and nitrates. The samples were also analysed for nitrites, ammonia nitrogen, total suspended solids, sulphates, arsenic and heavy metals (cadmium, copper, chromium, iron, manganese, mercury, nickel, lead and zinc), turbidity, pH and faecal coliform.

Electrical conductivity and pH were measured using the HANNA Instruments Models HI 98127 (EC) and HI 98312 (pH).

According to the water quality results (Table 2), pH values between 7.6 and 8.4 were measured in the rainwater samples collected from the six study sites. Of the various turbidity measurements for each of the samples, the E sample had the highest turbidity value, with a range of 2.34 NTU (Nephelometric Turbidity Unit), exceeding the standards of Resolution 2115 of 2007 that allows an NTU level in drinking water of up to 2.0 because a higher value may have adverse effects on human health.

No heavy metals were detected in the samples at elevated concentrations or at concentrations exceeding WHO standards. Concentrations ranging from <0.00019 to <0.15 mg/l were reported for cadmium, <0.05 to <0.20 mg/l for copper, <0.005 to <0.01 mg/l for lead and <0.05 and 0.31 mg/l for zinc.

Low concentrations ranging from <5 to 34.0 mg/l were reported for sulphates, with the highest sulphate concentration detected in sample D.

Manganese levels were analysed only for C because the detected manganese concentrations in laundry water can be affected by the process of washing jeans. The value reported for C was <0.15 mg/l.

Other parameters analysed were alkalinity and hardness. The reported total alkalinity values were between 14.3 and 34.0 mg/l, and the reported total hardness levels were between 12.0 and 54.0 mg/l.

Microbial contamination was also analysed; however, neither total coliform nor helminth eggs were detected in any of the samples.

The results of the Laboratory of Environmental Chemistry and Food analyses are presented in Table 2.

Rainwater pH normally ranges from 4.5 to 6.5, but the pH values measured for this study were between 7.6 and 8.4. These pH differences are explained by the results of studies conducted by various researchers. According to Göbel et al. (2007) and Meera & Ahammed (2006), the pH of rainwater may increase slightly after falling on the roof and during storage in tanks. These results are consistent with those obtained by Farreny et al. (2011), indicating that the pH of rainwater is in the range of 6.54–8.25 in several studies in Spain.

Other studies have found that increased pH levels in collected rainwater are due to the main components of the concrete used to construct the storage tanks such as Mg and orthophosphate o-PO4 (Wilbers et al. 2013). These compounds are dissolved when rainwater is stored in concrete storage tanks because the fresh rainwater is usually slightly acidic; the storage pH thus gradually increases as a result of leaching of these substances. Higher pH values of stored rainwater were observed in different studies, such as those by Lee et al. (2010), Lee et al. (2012) and Wilbers et al. (2013). The gradual leaching of these substances in storage tanks was also suggested by Villarreal & Dixon (2005) and Meera & Ahammed (2006), who observed the increase of pH evaluated during storage of water. According to the observations reported in the literature, most of the sampled collection systems, canals and reservoirs for buildings were built with concrete material, which can generate increased pH in captured rainwater.

Regarding the measured turbidity concentrations, high turbidity levels were only measured for building E, which had a value of 2.43 NTU. The WHO water quality standards indicate that turbidity levels up to 5 NTU are usually acceptable for water consumption. According to Resolution 2115 of 2007, a maximum of 2.0 NTU is allowed.

The high value of turbidity for building E could be explained by the presence of particles of matter that can come from the roofs and as a result of inadequate filtration or by the suspension of sediments in the distribution system.

According to Thomas & Greene (1993), Chang et al. (2004) and Torres et al. (2011), there is a strong relationship between the presence of industrial activities and the water quality of collected rainwater. For these authors, the presence of heavy metals such as Pb, Cu and Zn in samples of rainwater indicates that the air quality is poor and the presence of industrial areas, roads or highways near water uptake systems influence the detection of these heavy metals.

This study indicates that low concentrations of heavy metals were reported that do not exceed WHO standards. The sulphate concentrations for some samples (e.g., A (7.82 mg/l), D (34.0 mg/l) and B (33.0 mg/l)) are associated with the locations of these buildings close to highly congested vehicular routes. However, the reference standards (WHO, Decree 475 of 1998 and Resolution 2115 of 2007) allow for sulphate concentrations up to 250 mg/l in drinking water, indicating that the detected concentrations are low, when considering a household. Although the reported values are low, it is recommended that this water should not be used as potable water to prevent adverse human health effects, but it is considered safe water for other uses, such as irrigation, laundry and sanitary water.

Very low zinc concentrations were detected in the C sample (0.31 mg/l), but no source of zinc could be confirmed. The reported zinc levels may be a result of corrosion of brass fittings or downspouts. Previous studies have shown that galvanised metal roofs (Nicholson et al. 2009; Kingett Mitchell Ltd. 2003; Mendez et al. 2011) or painted galvanised metal roofing (Kingett Mitchell Ltd., 2003) produce high concentrations of zinc in the rainwater captured from these roofs.

Neither coliform nor helminth eggs were detected in any of the samples, perhaps due to the constant recirculation of water from high rainfall in the area, which can influence the reduction of microbial contamination of the stored water, particularly for buildings D and E. There is no evidence of the presence of animals (birds and cats).

The reported concentrations for the evaluated parameters were lower than the results reported in previous studies. One possible explanation is that sampling was conducted during the rainy season, and frequent rains would be expected to decrease the risk for accumulation of potential contaminants in the ceilings, such as leaves, sand, dust, and bird droppings, among others. Therefore, we can say that the physico-chemical quality of rainwater collected from roofs and reservoirs sampled at the six buildings is good and meets the water quality standards for domestic use. However, rainwater cannot be recommend for use as drinking water because there are no guarantees that the concentrations reported are maintained over time.

Rainwater is a potential source for domestic water use; however, the type of roofing material used for collection, and the proximity of the associated storage systems to highways and industries are factors that affect the water quality and influence the presence of contaminants.

The high rainfall in the suburbs of Pereira–Dosquebradas positively influences the quality of stored rainwater because its constant recirculation reduces the risk of accumulation of heavy metals, micro-organisms, and turbidity, among other contaminants. Therefore, the results obtained in this study indicate that rainwater collected from roofs is suitable for some household uses in the suburbs.

Owing to the untapped potential of rainwater in the suburbs of Pereira–Dosquebradas, an effort should be made to identify the possible uses of rainwater as an alternative source of water supply, thus reducing dependence on local and external sources.

Future research is expected to analyse the rainwater collected in seasons of El Niño, when a significant decrease in rainfall occurs and the rainwater sampling and analysis are conducted for stormwater storage tanks that are not washed or disinfected continuously.

The authors wish to thank the project ‘Metabolismo urbano y análisis ambiental del aprovechamiento de agua no convencional en edificaciones más sostenibles’ financed by the Vicerrectoría de Investigaciones, Innovación y Extensión de la Universidad Tecnológica de Pereira, Colombia.

1. Environmental Manager. Metropolitan Master of Environmental Management. Co-researcher Territorial Research Group Environmental Management, School of Environmental Sciences, Technological University of Pereira (UTP), Colombia. The Julita, Pereira, Risaralda, Colombia. A.A.: 097. migarcia@utp.edu.co.

2. Environmental Manager. Master in Operations Research and Statistics. PhD in Environmental Science and Technology. Investigator Research Group Territorial Environmental Management, School of Environmental Sciences, Technological University of Pereira (UTP), Colombia. The Julita, Pereira, Risaralda, Colombia. A.A.: 097. tito@utp.edu.co.

3. Environmental Manager. Student MSc in Environmental Sciences. Professor and Co-Investigator Research Group Territorial Environmental Management, Department of Environmental Science, Technological University of Pereira (UTP), Colombia. The Julita, Pereira, Risaralda, Colombia. A.A.: 097. mtflorez@utp.edu.co.