An innovative subterranean spring capture method for improved water quality

Water is an essential resource worldwide. A sufficient quantity of high-quality water is critical to the thriving of people. In 2016, 1.2 million deaths could have been prevented worldwide by improving access to clean water (World Health Organization 2019). These deaths are largely attributed to pathogens found in unclean water. Diarrhea, the second leading cause of death for children under five, can be caused by Escherichia coli, which is a common pathogen in untreated water (World Health Organization 2019).

For rural communities in poverty, health issues related to water quality are amplified. In these communities, sanitation is undervalued due to a lack of education, and households have insufficient funds to pay for improved sanitation practices. As a result, infection rates of water-borne diseases are higher. In Ecuador, our country of focus, these recurring infections keep rural community members trapped in poverty, as any extra income is directed to medical bills and caring for the sick (Leaska 2019). Therefore, clean water is essential to both the health and financial well-being of these rural populations.

The WHO identifies microbial contamination as the prime risk affecting the viability of rural water systems. In Ecuador, the importance of rural water systems is plainly observed: 75% of the urban population in Ecuador has access to safely managed drinking-water services, while only 53% of rural Ecuadorians experience such access. This disparity is significant because rural community water supplies serve 36% of the country's 16 million inhabitants (World Bank 2018). In order to be classified as ‘safely managed drinking-water’, the service must have an improved source that is accessible on premises, available when needed, and free from fecal and priority chemical contamination. The engineered strategy to deliver safe water includes improved water sources such as piped water, boreholes, or tube wells, protected dug wells, protected springs, rainwater, and packaged or delivered water (World Health Organization 2021).

Ecuador is home to many natural, high-quality springs. Two types of springs, alluvial and topographical bedrock, are commonly observed within the central mountainous portion of the country, each with its own design challenges and potential for contamination. The alluvial springs can be characterized as groundwater flowing downgradient through alluvium in a valley. These springs are vulnerable to contamination from latrines and nearby grazing animals. Topographical bedrock springs are good indicators of the local piezometric surface elevation in the aquifer that results in water flowing out of the side of a hill or through a minimal distance of colluvium. Because the recharge point for these springs is typically very far from the spring eye, water from bedrock springs tends to be older with less risk of contamination than alluvial springs. However, due to the steep terrain surrounding springs, there is potential for landslides, and access for construction is a challenge.

Regardless of the type of spring, protecting the spring from contamination is a step toward delivering risk-free water. Springs have traditionally been protected using a variety of methods including structures built above ground, like spring boxes, or underground systems, like qanats, and subterranean captures, also called seep fields. Most of the springs in the mountainous portion of Ecuador are either spring boxes or subterranean captures.

While the government sets a standard of practice, it is left up to local efforts to deliver high-quality water. Since the formation of the United Nations Sustainable Development Goals in 2015, non-governmental organizations (NGOs) across the globe have made a concerted effort at improving water quality. In Ecuador, there is a joint effort between NGOs and local communities to design, install, and maintain rural water systems which have led to a democratization of water access and resource management. These various stakeholders can address the needs of a local community in a greater manner than the centralized government is able. However, decisions must be made specific to a locality that is embraced by the local community for the ultimate success of the water system. In fact, it will be the engineering decisions that will determine the starting point of a rural water system's success. If a system can capture water free of contamination and prevent recontamination from surface and rainwater, the maintenance and management of the system becomes simple and affordable while also distributing high-quality water.

One example of such an organization is Corporación de Desarrollo Integral Socio Económico (CODEINSE), a long-term local Christian ministry and nonprofit organization founded in Ecuador that has developed an improved method of spring capture that decreases the risk of water contamination while improving spring yields and being economically sustainable. This new technique, ‘the CODEINSE method’, builds on the strengths of traditional subterranean spring captures while employing several design improvements like a concrete cap and sufficient gravel pack to prevent surface water contamination.

Using water quality data collected in Ecuador from 2020 to 2021, a comparative analysis of CODEINSE's method and other common methods of gravity-based spring capture will be presented. This work surveyed a variety of water capture methods in an effort to critically evaluate the effectiveness of eliminating microbial contamination at the source.

This study was carried out in the area surrounding Guamote in the Chimborazo province of central Ecuador. Approximately 240 km south of the equator and positioned in the Andes Mountains, the area experiences a mostly dry and temperate climate, with an average annual temperature of 15 °C and an annual total precipitation of 460 mm, consistent with a tropical alpine climate. Eighteen communities were selected for study displaying a variety of water capture strategies and designs. Each community had a small population of between 200 and 1,500 residents.

All the water distribution systems were either gravity fed or a combined pumped and gravity-fed pipe system originating from either a groundwater source or surface water. The groundwater sources employed different spring capture methods. It is common for more than one spring capture point to feed into a water distribution system. As communities grow, sites may be added to increase supply. As a result, a total of 41 different water capture sites from 18 communities were investigated.

Each water collection source went through a series of assessments. Initially, the source water was classified as the type of influent spring water, such as alluvial or topographic bedrock. Furthermore, the catchment area was assessed for spring protection, land use, and potential contamination sources. The engineered spring capture design was then categorized into four categories: spring box, subterranean capture, CODIENSE, or surface water. The categorized spring capture design was queried for the structural durability and evidence of ongoing maintenance.

Spring box capture is one of the most widely adopted spring capture methods in Ecuador. A spring box is typically a poured-in-place concrete box that surrounds a spring eye and protects it from contaminants such as surface runoff or rainwater. A typical installation is shown in Figure 1.

Figure 2 displays a typical installation of a subterranean spring capture. Subterranean spring capture is a method of spring capture that utilizes the filtration capabilities of the surrounding soil and added gravel before collecting the water in a tank. In a subterranean spring capture, trenches are dug in a field with shallow groundwater. A perforated pipe is placed in these trenches, surrounded by gravel, and then covered with backfill (Jones 2014; Skinner & Shaw 2017).

CODEINSE has developed a particular spring capture design displayed in Figures 3 and 4. As such, the CODEINSE model spring capture is broken out separately from the others. This model is most similar to a subterranean capture, using filtration through soil and gravel, but has unique features including the use of plastic sheeting, geotextile fabric, a concrete cut-off wall, as well as a concrete cap to protect spring water from outside contamination.

In the geographic area of study, there are existing regional water systems. Episodically, these systems have tapped into surface water, typically streams. While no community was actively using surface water as a drinking water source, it serves as a comparable alternative if no improved water is available.

Bacteria testing was performed using Easygel Coliscan bottles from Micrology Laboratories. Typically, 1–5 mL of raw water was collected using serological pipettes from the source and mixed with media. The standard draw was 5 mL except when an elevated level of contamination was suspected. Under high microbial presence, 1–3 mL was drawn to allow for reasonable colony growth. The mixture was plated on an easy set Petri dish and set within 30 min. The plates were allowed to incubate at room temperature for 48–55 h, resulting in fully developed colonies. A colorimetric indicator was released upon the microbial activity affording the differentiation of coliforms and fecal coliforms. Fecal and non-fecal coliform counts are normalized and reported as colony-forming units (CFUs) per 100 mL of sample water.

There are many factors that contribute to a safe and effective water system. The WHO has established a water safety plan for community water supply (World Health Organization 2017). The Peace Corps has a WASH matrix for ranking all aspects of a water system (Galicia 2019). Of the many aspects that contribute to a water system's quality, the engineered and constructed point of collection is essential and permanent. Therefore, the analysis focused on the spring capture method as a critical component to delivering safe drinking water. The WHO standard for high-quality water is defined as the absence of CFUs, essentially being free of microorganisms (World Health Organization 2017). However, the presence of microorganisms in water is not a uniform level of risk. The assessment of microbial risk to a water system is quantified according to the following classification (Table 1) from the WHO, where an increasing level of microbial contamination is associated with an increasing risk assessment level.

While visiting spring capture sites in Ecuador, surface bacteriological tests were conducted on the interior surfaces of the collection boxes, typically an interior wall and the underside of the access hatch, using sterile cotton swabs. After swabbing the area of approximately 120 cm², swabs were then immersed in the Coliscan medium to begin the bacteriological culturing process. Plating of the medium was carried out with the same procedure as that of the bacterial testing.

The site investigation and data collection occurred over several months between January 2020 and August 2021. Each site was investigated for land use, potential contamination sources, and spring capture structure integrity. Interviews with the local Aguatera were conducted to glean more information about spring yield, construction practices, and maintenance demands. Bacterial testing was performed at springs and access points along the water distribution line followed by an analysis of the data. Figure 5 shows the general methodology procedure.

Microbial contamination was determined for fecal and non-fecal coliforms from the 41 studied water capture sites. Fecal coliforms, which include E. coli, have been linked to disease manifestations and are considered an assessment of risk, as they represent the inclusion of intestinal originating bacteria from warm-blooded organisms. However, the long-term health of a water system can be assessed from the current level of total coliforms, a summation of both fecal and non-fecal coliform. This is reflected in the WHO standard classification (Table 1). This gives a measure of the system's potential to host the growth and development of pathogenic organisms into the future. Notably, the water capture sites in this study showed ranges of 0–2,000+ fecal coliforms and 0 to 5,000+ for total coliforms. This range represents the full spectrum of WHO risk categories. The collected data are listed in Table 2 and sites are geographically displayed in Figure 6, with their risk assessment and spring capture type labeled.

A spring box is one of the most widely adopted spring capture methods in which a poured-in-place concrete box surrounds a spring eye in an attempt to protect it from contaminants such as surface runoff or rainwater. A representative installation is shown in Figure 1. Two key benefits of a spring box capture are that it is low cost and low maintenance (Hart 2003). However, they are also subject to risks associated with poor construction and design vulnerabilities. These vulnerabilities include access points, the overflow pipe, and limitations from their assumption that water production is localized rather than dispersed. A diversion trench constructed around the box, as shown in Figure 1, could help direct surface water away from the box, but these can accumulate sediment over time and are often not implemented.

At two Department of Water Affairs and Forestry spring protection locations in South Africa, the spring boxes did not fully enclose the spring eye and the spring moved to another location (Lenehan 1996). The risks associated with missing or not fully enclosing the spring eye are ponded water, reduced yield, and eventual failure. The ponded water transforms groundwater into surface water, which is ripe for microbial growth. Spring boxes can be particularly vulnerable to surface water through their overflow pipe, hatches, and construction defects, as they would allow surface water to permeate into the box. Figure 7, a spring box in Cebollar, shows one such example of this. There is ponded surface water surrounding the box. Water tested from within this box had a total coliform count of 2,060 CFUs/100 mL.

Our data showed the susceptibility of spring boxes to contamination. Of all the capture types investigated, spring boxes showed the most variance in coliform levels, 0–3,107 CFUs/100 mL. Figure 8 shows the coliform counts sorted in increasing percentiles by water type into relative risk levels from the WHO categories. Surface water has the highest level of coliform counts. Though it is rarely used for drinking water, it is included as a reference for unprotected water. At the 25th percentile, spring boxes showed only 7 CFUs/100 mL, a low risk, while surface water yielded 1,400, already a high risk. However, at the 85th percentile, spring boxes had almost 3,000 CFUs/100 mL to surface water's 5,000. In fact, spring boxes were a very high coliform risk at more than 25% of surveyed locations.

The subterranean spring capture, as seen in Figure 2, utilizes the filtration capabilities of the surrounding soil and added gravel before collecting the water in a tank. The main benefits of a subterranean spring capture system are that it is simple to construct, efficient in capturing spring water, and able to protect the spring even as the geology shifts over time (Jones 2014). Subterranean spring captures can stretch over a large area, allowing for much water to be collected. Similar to spring boxes, a key design feature is to provide the least restrictive path for water to flow into the tank. The gravel pack and pipe provide an easier path for water flow than the adjacent soil, resulting in consistent spring water collection to a downstream water collection tank.

However, an inherent risk to the subterranean spring capture system is the vulnerability to surface water contamination, depending on the depth of the perforated pipes. If they are installed in a shallow trench, surface water may enter the system without thorough soil filtration. Furthermore, subterranean capture is subject to increased and difficult maintenance because of clogging pipes (Hart 2003). Sediment travels with the water through the gravel resulting in sedimentation in the tank and clogging of pipes. The perforated pipes must enter a local collection point which shares common concerns to spring boxes, such as access points in the tank and overflow pipes that provide a possible means of contamination from surface or rainwater.

The coliform counts from subterranean captures yielded values that ranged from 0 to 1,740 fecal and total coliforms (Figure 8). At the 25th percentile, subterranean capture was already an intermediate risk, as there were 32 CFUs/100 mL. In comparing subterranean capture to spring boxes, there is less variance among the subterranean capture. By the 75th percentile, the coliform counts have grown to only 700 CFUs/100 mL, far less than what is seen in spring boxes at 2,000 CFUs/100 mL. The median values for both capture methods are nearly the same, 90 versus 110 CFUs/100 mL for spring boxes and subterranean, respectively. At the site shown in Figure 9, total coliforms were measured to be 1,740 CFU/100 mL. Consequently, both methods typically yield an intermediate to high level of microbial risk. The spring box may be an effective spring capture method, but it is highly irregular. While the subterranean capture is less variable, it still leaves much to be desired in regard to consistent water quality and quantity.

The ‘CODEINSE method’, developed over the past few decades, builds on the strengths of a traditional subterranean spring capture while addressing some of the major drawbacks. Like subterranean capture, the CODEINSE method takes place underground; however, the CODEINSE method employs several design improvements over traditional subterranean spring capture to address both long-term maintenance and surface water contamination concerns.

The CODEINSE design draws on the building blocks of the subterranean spring capture by utilizing the natural soil and gravel filtration while limiting the spring water's air and surface exposure. In addition, the design inherently reduces maintenance requirements and results in community benefits. To minimize maintenance, the CODEINSE method uses geotextile, plastic sheeting, and washed gravel to control and reduce the flow velocity entering the system, thereby limiting the movement of fine granular material, and preventing or reducing sedimentation in downstream tanks. For example, in Ulpan Ainche, when CODEINSE replaced a subterranean spring capture, the community saw a substantial improvement to their sedimentation issues and water quality. While these effects were not quantified at the time, as a result of cleaner, higher quality water, the community dairy farmers were able to sell their milk at a higher grade than previously attainable. Furthermore, the CODIENSE method looks to minimize microbial contamination during and after the construction of the spring capture. The installation of the gravel pack is followed by a sodium hypochlorite shock to kill the associated surface contaminants introduced via the construction materials.

After the system undergoes the sterilizing shock, plastic sheeting is laid over the top of the gravel pack and a concrete cap is poured to provide a protective barrier from the contaminant carrying surface water runoff. The CODEINSE method uses a concrete cut-off wall surrounding the downstream portion of the spring eye to prevent water from bypassing the gravel pack and locates the tank further downstream than a traditional subterranean system. This combination ensures adequate pressure from the spring to the tank and reduces the probability of spring capture failure. Additional practices such as reducing the points of exposure to the atmosphere and using secure access covers to reduce contamination from surface and rainwater are employed. A CODEINSE spring capture can be viewed in Figure 10.

The CODIENSE method has an assortment of interventions to minimize surface exposure, and our data reflect a significant improvement in the captured water. The variance between the 25th and 85th percentile ranks of the CODIENSE sites is the least of all measured spring capture categories, ranging from 0 to 95 CFUs/100 mL (Figure 8). In fact, over 50% of CODEINSE spring captures tested showed 0 total coliforms, placing it into the no risk category, making it ready for human consumption with no further treatment. At the 75th percentile, the CODIENSE collected water shows between 15 and 50 times fewer coliforms than the subterranean and spring box waters.

In order to determine the significance of the apparent CODIENSE method reduction in microbial presence, a χ² analysis was performed. To maximize the level of statistical robustness, the WHO categories were grouped. Specifically, the ‘No Risk’ and ‘Low Risk’ categories were combined and are referenced as a low-risk group. The ‘Intermediate Risk’, ‘High Risk’, and ‘Very High Risk’ categories were combined and are referred to as the high-risk group. This grouping was chosen because the WHO considers the no risk category to be ideal but the low-risk category to still be acceptable for human use. As a result, the combined categories reflect a minimal risk, and consequently, usable water versus water in need of treatment to become usable. The spring capture methods are grouped based on the observed differences within Figure 11. The CODEINSE method is compared against the combination of spring box and subterranean capture. In doing this, there is sufficient data in all categories to provide a meaningful comparison. Data counts for the two-by-two χ² analysis are shown in Table 3. A two-by-two χ² analysis shows that the distribution of low- and high-risk water is significantly different at a 99% confidence level. In essence, the CODEINSE method shows a sizable increase in the low-risk category which is interpreted as giving a superior method of groundwater capture.

Groundwater is generally considered to be microbially ‘pure’ (Pedley & Howard 1997). However, the water described thus far has shown a variable and at times large presence of microorganisms. This suggests that the groundwater is being reinfected from some intruding source. At various spring capture sites, spring box, subterranean, and the CODEINSE, the surfaces of the spring box or collection box were tested with sterile swabs to examine if there was a measurable difference in surface coliform counts. All capture types showed a range of surface coliform presence, from 0 to 0.89 CFU/cm². A biofilm community might be washed into the water system by rainwater. However, there was no statistical difference in the level of coliforms measured from the concrete surfaces of any spring collection site. As a result, the reinfection of the water systems seems to arise from the intrusion of surface water. At all measured locations, surface water displayed a very high risk from coliform counts. The intermingling of surface and collected spring water could account for the increasingly high counts observed in the spring box and subterranean capture sites.

The spring box, subterranean, and CODEINSE methods were assessed in the field to determine design features that may contribute to observe microbial trends and risks associated therein. Surface water is distinct from groundwater and shows a higher level of microbial presence. Any of the engineering methods investigated yield improved water quality because groundwater is captured before it is surface exposed. However, the CODEINSE method provides a statistically significant increase in water quality over other capture methods. The observed improvement may stem from the intended design aimed at excluding surface water and surface exposure both during construction and throughout spring water collection.

In conclusion, the CODEINSE spring capture clearly presents an improvement on traditional methods of spring capture. Through implementing an uncomplicated, yet effective design, a CODEINSE spring capture not only provides better quality water on average, but it is also simple enough that a committed, nontechnical population can construct their own spring captures in the future. The resilience of the CODEINSE spring capture to poor construction makes it highly attractive for implementation in areas of the world where skilled construction labor is scarce, or too expensive to hire. For more direction on how a CODEINSE spring collector is constructed, the Clean Water Institute at Calvin University is producing a design manual in tandem with CODEINSE that should be available by Summer 2022. By maintaining the quality of groundwater and preventing contamination, the CODEINSE method could be utilized in locations where sources of clean groundwater are available. The CODEINSE method has already provided cleaner water to the rural Andean communities in Ecuador, but it has the potential to do so in communities in developing countries across the world. With improved access to clean drinking water, not only will these communities be healthier, but they will also be enabled to escape poverty and improve their overall quality of life.

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