A low cost, reusable, sustainable, batch water purification system

The World Health Organization and the United Nations estimate that approximately 800 million people lack potable drinking water (Scott 2013; VOA 2009.). While this problem is largely worldwide, the problem is especially acute in developing countries were lack of potable water is a great challenge. These facts are well known and non-profit organizations, governments, and for-profit companies are addressing the problem through a number of initiatives (Scott). Nevertheless, there is a need for additional technologies and combinations of technologies to reduce the severity of this crisis.

The problem can be exacerbated in emergencies. Lantagne and Clasen (Lantagne & Clasen 2012) studied the effectiveness of household water treatment in four acute emergency situations. In three of the situations, only chlorine methods of purification were used, however, in the fourth situation, chlorine was used in conjunction with a filtration technique. The ‘effectiveness’ (see their definition of effectiveness) ranged from 0 to 67.5% ‘with only one pre-existing chlorine program in Haiti and unpromoted boiling use in Indonesia reaching >20%’. They found that familiarity and training with the technique and its materials was a key to success in such situations and suggested metrics for improving and evaluating the effectiveness.

Hence, there is often a continuous problem and an acute problem in the same geographic area. Even though it is imperative that capital costs and operating costs be as low as possible, there are a number of technologies that are viable. Technologies that fulfill these constraints include chlorine tablets, boiling, clear plastic bottles exposed to UV light from the sun, LifeStraws (LifeStraws) that filter water on an individual scale, and hand dug wells. However, all of these technologies have weaknesses. Chlorine tablets need to be supplied and distributed by an outside source. Boiling requires energy, often from wood, that is in many situations scarce. Clear plastic bottles and (LifeStraw, 2014) need to be supplied. LifeStraws are rated to filter up to 18,000 l of water; however, their use of a filter means they are susceptible to clogging by particulates. Finally, wells provide a tremendous source of potable water provided the well water is close enough to the surface, potable, and of sufficient volume which is not always the case.

The use of metals as an antimicrobial is well known and has been used for millennia (Landau 2007). Recently, there has been renewed interest in using metals to kill microbes (see for example: Thurman & Gerba 1989; Abushelaibi 2005; Michels Noyce & Keevil 2009; Shrestha et al. 2009; Sudha et al. 2009; Copper Development Association 2010; Grass et al. 2011; Zhu et al. 2012; Leite & Padoveze 2012; Bindhu & Umadevi 2014). There is clear evidence copper and silver is effective (Michels, Zhu, Leite) as an antimicrobial surface. There is evidence that silver and brass surfaces are effective (Shrestha). Gold nanoparticles are also effective as antimicrobials (Bindhu). As one looks to a material for use in this technology, gold is very expensive, and there are negative health effects associated with colloidal silver and nanoparticles of silver (Hadrup & Lam 2014). Thus these materials are not good choices in terms of cost and safety. Moreover, there are no metabolic processes dependent on silver (Lansdown 2007).

This suggests copper as the metal of choice. The World Health organization in 2004 (WHO 2004) recommended a maximum copper concentration of 2 mg/l for water and a maximum tolerable intake of 10 mg/day. They also note that at levels above 2.5 mg/l, copper imparts a bitter taste to water so its presence in water should be noticeable. Furthermore, copper is an essential trace element for the human body. Trace elements in river water were investigated by Sholkovitz and Copland (Sholkovitz & Copland 1980). The solubility of Cu was found to be near or below 20 μg/l when the pH is between 9.5 and 4. Thus, except in cases of metabolic disorders of copper homeostasis (WHO), copper in water should have small or negligible adverse health effects.

Since the situation is most acute in the poorest countries, low-cost and low-maintenance, easy to use, and locally sustainable technologies with minimal, necessary, external supplies that may be disrupted would be of interest to those addressing these issues. In this brief paper, we present the results of a study that greatly reduced, and in many cases eliminated coliforms in water by agitating contaminated water in a bottle containing copper foil and/or screen. We lacked the means to extend the study to other microorganisms such as viruses; however, Thurman and Greba (Thurman & Gerba 1989) indicate that this method is effective against some viruses, and in conjunction with other technologies, this technology may be a part of the solution to this vexing and continuing problem.

The goal of the project was to create a technology that is simple, low-cost, reusable, has low energy consumption, and uses readily available materials. Consequently, copper and silver nanoparticles were immediately eliminated because they are not readily available in developing countries without an external supplier. Lantagne and Clasen suggest that the technology should be intuitive to use, simple enough that a child can purify water, and sustainable so that the technology does not require extensive outside support that may be cut off in times of conflict and emergencies. We believe our process is scalable to some extent, but tests have been limited to water volumes between 0.1 and 0.8 l.

The use of copper in some manner is consistent with our goals as summarized in the introduction. This section describes the experimental details.

The polyethylene bottles and glass bottles were obtained from Fisher Scientific, USA. The copper screen and foil came from multiple sources including foil found in the lab, Basic Copper (Basic Copper, 2013) and Online Metals (Online Metals, 2014), the screen, also known as copper wire mesh was obtained from on line sources. All copper was commercially pure (99+ %). Attempts to find less pure copper without jumping to 90% copper 10% zinc were not successful. The foil was 25.4 μm thick. The water was obtained from a polluted, local stream so that the experiment would represent a real-life situation. The microorganism load of the water varies throughout the experiment, but each test was compared to a control. The testing media for coliforms was Coliscan Easygel^TM from Micrology Laboratories (Microlgy labs, 2015). This test is simple and effective for water testing. The copper content of the purified water was less than 0.5 mg/l (indistinguishable between 0.0 or 0.5 mg/l, the lowest calibration points for the test). This was determined by using test strips (SenSafe, copper check, 2015).

The procedure was straightforward. Water was collected in a 4 l polyethylene bottle and used within 4–6 hours. The volume of water used in a given experiment was determined using a graduated cylinder. In all experiments, for comparison purposes, an identical, empty (no copper) bottle was filled with the same amount of water from the same source. This bottle served as a reference for the effectiveness of the technology. Depending on the experiment, one or more bottles containing a known volume of water and a known weight of copper where prepared. The ratio (mass of water)/(mass of copper) was used as a variable as it led to better correlation to the results than volume of water to estimated surface area of copper. The conversion from mass to volume assumed the density of water was 1000 kg/m³. The test bottle was agitated for a time determined by experiment design and measured with a timer. After the designated agitation time was complete, 0.002 kg of water was removed from both the test bottle and the reference bottle, mixed with the EasyGel chemicals, vigorously shaken and added to a labeled Petri dish with growth media that is also part of the Easygel system. These Petri dishes were kept at room temperature (25 ^oC) for 2 days (per Easygel directions) and then visually examined. At this time, pictures were taken of the Petri dishes for later study. When longer shaking time on the same sample was desired, the test bottle of water and copper was agitated again for a set time and an additional 0.002 kg of water was removed and the test procedure described above was repeated. In order to replicate real world circumstances, bottles and copper were reused without washing.

This section presents the salient findings. Details of colony forming units (CFUs) for various water–copper ratios and agitation time are not graphed because the data depend on the initial (variable) water quality. A drawing of the apparatus is shown in Figure 1. We found that when a water–air–copper interface was present, the results were significantly better and faster than when using a completely full bottle.

The results of the first experiment, shown in Figure 2, indicate that agitation is necessary for this technology to work. Figure 2 shows a photo of two petri dishes after 1 day, including a control sample on the left which contain water that was not exposed to copper and the test sample on the right which contains water that was exposed to copper. Neither sample was agitated. The number of CFU per liter is 500 times the number of red spots observed in the Petri dish. The control and the non-agitated experiments show a similar number of CFU. Thus the water in this test had 50,000 to 100,000 or more per litre.

Once it was determined that a water to copper mass of 10:1 worked well, experiments with aggregation times between 15 minutes and 2 hours were undertaken. The authors decided that a thirsty person is unlikely to wait for more than 30 minutes to drink the water, and thus added the objective that the purification occurs within ½ hour. Keeping the water-to-copper ratio at 10:1 and limiting agitation to ½ hour, results such as shown in Figure 3 were obtained. In this experiment, copper foil (25.4 μm*15.24 cm*43.2 cm) and copper screen of the same dimension were used. In the experiment shown in Figure 3, the foil and bottle had been used previously and were shook dry and exposed to air for a few hours before the experiment.

The results in Figure 4 illustrate the results of an experiment that used a 500 ml sample of contaminated water. The sample on the right is the unpurified water and shows at least nine CFU (red dots) obtained from a 2 ml sample of microbial contaminated water. This corresponds to roughly 450 CFU/100 ml. The sample on the left is from the purification device after shaking and sitting and shows no CFU. Testing of this water even after 3 weeks in the container with the copper shows copper levels near 0 mg/l.

These are typical results and have been reproduced many times. Additional tests were carried out for over a year with the intension of improving the process by reducing the necessary agitation time and/or increasing the water-to-copper ratio. Other variations and geometries were also explored. None of the additional tests resulted in an improvement to the process and have not been reported or pursed. Since the plastic bottles used could transmit UV light from the lab's florescent lighting, the experiment was repeated with glass bottles and similar results were obtained.

It was realized that in real-life applications, the water may contain suspended particles and large-scale impurities such as sticks, leaves and the like. To simulate this situation, turbid water was collected and a funnel was made of copper and covered with an old piece of cloth. The water was poured into the apparatus through this crude filtration system. The results were identical to those obtained with significantly less turbid water. In fact, even when there was mud at the bottom of the bottle after letting the agitated sample sit for several hours so the sediment could settle, the water experienced a significant decrease in microbial contamination.

This technology is simple and results are reproducible as long as there is sufficient time for agitation and the water mass to copper mass is ≥10:1. This approach does not address chemical contamination of water and has not been tested for all types of bacteria and viruses. Nevertheless, these results indicate that when used separately or with other technologies this technology has the potential to be useful. For example, Lata and Samadder (Lata & Samadder 2014) review removal of heavy metals from water using rice husks. Since rice husks have little commercial value and even untreated husks have some effectiveness, using rice husks as a filter media has the advantage of removing both heavy metals and particulates. Although, it is not clear from their review how long adsorption of heavy metals takes, it may be possible to combine rice husks and copper in the same container for removal of both types of impurities. These authors also point out that husks treated with various chemicals are more effective absorbers of heavy metals than untreated husks.

Other biological absorbers have also been examined by others. For example, Chaparadza and Hossenlopp (Chaparadza & Hossenlopp 2012) used processed banana peels to remove atrazine (a commonly used weed killer) from water. While this contaminant is probably not a major problem in developing countries, the use of agricultural waste as a filter and absorber of heavy metals in water is added to this technology or even inside the container may allow both heavy metal and microbial contamination to be removed using locally available materials.

It is interesting that our results are substantially different from those of Sudha et al.. They found that when water was stored overnight in a copper pot or a glass bottle with a copper coil water that had 500–1,000 CFU/ml was purified so no CFU could be grown. Our results (Figure 2) show that without agitation little purification occurs. This suggests that the standard copper pots purchased at a kitchenware shop and cleaned using tamarind and salt that Sudha et al. used might have extremely useful trace impurities. The copper in our experiments was of purity >99 + %, and we avoided cleaning it to simulate worst case situations. We were unable to procure copper of known lower purity than commercially pure. The use of tamarind to clean copper and other types of ‘copper’ should be investigated.

This technology has the potential to support local/regional businesses making containers, copper foil from billets and assembling the completed product. This means that organizations supporting this effort can focus on supplying relatively inexpensive raw materials rather completed devices.

Finally, by using a careful chosen geometry, this process may be enhanced when exposed UV radiation from the sun.

This work was supported by Kent State University and was an undergraduate research project completed by D. Enders. We thank W. T. Southards and J. Ruller for useful discussions.