Effectiveness of point-of-use (POU) filter system for removal of contaminants from water

Groundwater supplies most of the drinking water in the small settlements in the Danubian plain region of Bulgaria. The region is mostly agricultural, leading to chemical pollution, mostly with nitrogen, of the shallow groundwater. Direct health concerns, linked to nitrates contamination of water, determine the increased popularity of drinking water point-of-use (POU) filtration pitchers. The objective of this study was to evaluate the effectiveness of a commercial pitcher filtration system in removing select pollutants, such as metal ions, namely, copper and iron, nitrate nitrogen, and suspended solids (algal culture of Monoraphidium contortum) from augmented tap water. In addition, we have studied the effects on electrical conductivity and pH of water. The POU system of choice was a widely used brand name in Bulgaria, providing cartridges that fit most of the European pitcher filter brands. We have evaluated the performance of the filters at three different exploitation intervals – pre-washed new filter (0%), 50%, and 100% of the exploitation capacity set according to manufacturers’ claims of recommended volume of water, as well as at different contact times between the water and the filter media. The results indicate that the efficiency of the filters diminishes with aging and increases with increasing contact time – multiple filtrations. The efficiency in terms of electrical conductivity and amount of iron decreases proportionally with filter age, and in terms of phosphates, the maximum effect is observed in filters at 50% capacity. Water filtration reduced water conductivity by 12% in a single filtration and stabilized the pH values towards the neutral range. The effect on pH values is inversely proportional to the buffering capacity of the water. The filters removed 85–98% of copper ions and 20% of iron ions. Nitrate removal efficiency averages 40% and doubling the contact time increases the efficiency to 70%. The efficiency of removing suspended solids is on average 17%, mainly due to the small size of the particles. Overall, the POU systems are an effective way to purify water at home. The filter cartridges effectively reduce contaminants, such as metals and nitrates, which are particularly problematic in areas supplied by groundwater sources.


Introduction
Apart from the availability of water, its safety is a top priority and a challenge for the existence of modern society [1].Due to its chemical properties, water is often considered the "universal solvent", so it can easily get enriched with a variety of substances, of natural or anthropogenic origin.Water naturally cleans itself via self-purification, filtration through the ground, and evaporation via the water cycle, but the continuous disposal of wastes contaminates the waterways and may compromise its quality.Providing settlements with drinking water that meets health requirements according to modern regulatory documents is becoming an increasingly difficult task.This is due to the pollution of natural waters used for drinking water supply [2].
The term potable water refers to water for drinking and domestic purposes, regardless of its origin and whether it is supplied through a centralized water distribution network, or by any other means such as tankers, bottles, boxes, etc [3].Potable water is usually obtained from surface (springs, rivers, lakes, and dams) or underground sources and is considered safe when it does not contain 1305 (2024) 012012 IOP Publishing doi:10.1088/1755-1315/1305/1/012012 2 microorganisms, parasites, chemical, radioactive, and other substances in number or concentration that represent a potential danger to human health [3].
According to Bulgarian legislation, the water supply operators are responsible for the quality of drinking water supplied to the population, as well as for the maintenance and operation of water supply systems and the provision of water supply services to the designated territories [4,5].In recent decades, due to increased migration, the population of the smaller settlements in parts of the country has been decreasing, which slows down investments in infrastructure [2].Thus, water supply operators are facing very serious problems and challenges: use of outdated and depreciated water supply facilities; water sources located in unsuitable places in arable agricultural land; increased content of nitrates in groundwater used for drinking water supply; lack of established sanitary protection zones around the water sources.According to Arlosoroff [6], shallow groundwater sources, like wells, are most susceptible to contamination by the adjacent territories.
Due to periodically occurring or permanently existing problems in the drinking water supply in the country, interest in devices and systems for decentralized additional purification of drinking water (point-of-use filters) has increased in recent years [7].They are used to optimize tap water concerning the fact that some people are in search of better-tasting water, while others look for a way to avoid health concerns related to tap water consumption [8].Addressing the quality of drinking water directly at the point-of-use is considered economically beneficial, given that only 3% of the total daily water consumption per person is used for drinking.The main element of point-of-use jugs is the filter cartridge containing a mixture of sorption materials, which can absorb certain impurities from the water.The point-of-use systems usually combine several principles of purification, such as mechanical filtration for removing physical impurities such as fine sand, detached pipe scale, silt, and plankton found in the water; activated carbon filters; ion exchange resins; filters using the reverse osmosis process and ultraviolet light [9].Although most point-of-use filters have similar composition and arrangement, the selection of the proper filter should be based on the water quality and the specific pollutants in the particular water supply network [10].Despite Maletsky's conclusions that point-of-use filter pitchers represent the most cost-effective system for domestic drinking water treatment [11], previous studies indicate that the majority of these pitchers may fall short of achieving water quality standards, and that advertising information provided by manufacturers is not consistently accurate [12,13,14].
The purpose of this study is to determine the effect of the exploitation time of the point-of-use filters, and the contact time of the water and the filter loading (media) on the efficiency of water purification.To achieve this goal the following tasks were performed: 1. Study the influence of exploitation intervals (filter age) on the efficiency of water purification from suspended solids, metal ions and nitrates/phosphates. 2. Investigate the effect of the contact time between the water and the filter media on the removal efficiency of major impurities from the model water through sequential filtration.

Materials and methods
During the experiment, we tested single-brand sorption-type filters for their efficiency in the purification of water from impurities.While some general information was available from the filter manufacturers, the internal composition of the pitcher filter media was further investigated after the dismantlement of the devices.The cartridges contained a mixture of exchange resins, which visually appeared to be comprised of two components (differing in size and colour), and coconut shell activated carbon in a ratio of approximately 1:1.5 (w/w), evenly distributed in a fibrous material for prevention of compaction and maintenance of a uniform water flow through the filter.As a means of preventing the removal of filter material, the cartridges contain polypropylene mesh screens, at their top and bottom, with an eye opening of 80 μm.The water flow rate of the water through the cartridges was on average 220 ml/min (200-230 ml/min) and, according to the manufacturer, the service life of the cartridges is 200 litres.
The experiments were based on filtering of the model water samples through filter pitchers at three different exploitation intervals (filter age)pre-washed new filter (0%), 50% (100 litres), and 100% (200 litres) of the exploitation capacity set according to manufacturers' claims of recommended volume of water.Additionally, we investigated the effect of the contact time between the water and the filter media.To address this issue, we have made 10 successive filtrations of the same model water sample through the filters.Water (50ml) was taken for analysis after 1 st , 2 nd , 3 rd , 5 th , 7 th , and 10 th filtration, corresponding to up to ten times increase in the contact time between the water and filter media.The sequential filtration was performed with each set of model water on different filters.Thus, we used a set of filters for the metal retention analysis, and other sets for the analysis of the nutrients and suspended solids retention efficiency.Simultaneously to the retention analyses, we have measured the pH and electrical conductivity of the water samples.
In the present work we tested the purification efficiency by the following indicators: -Suspended solids (SS).Assessed by the retention of the alga Monorafidium contortum (Thuret) Komárková Legnerová 1969 grown in laboratory conditions.The retention efficiency was determined by the change in the amount of Chlorophyll-a in the phytoplankton, according to the international standard ISO 10260.-Metal ionscopper and iron.The amount of copper and iron in the water was determined spectrophotometrically on WTW Photo Lab 7100Vis, by the Spectroquant method 14767 and 14761 of Merck.Copper was selected as a representative of contamination of drinking water by heavy metals.Since the concentration of copper in tap water is below the detection levels of the method (0.01 mg/l), a model water was created with a copper concentration of 4.5 mg/l.As a source of copper, we used copper sulphate.The model water was prepared immediately before the experiment, from a stock solution with a concentration of 1000 mg/l.Dechlorinated tap water (aged 48 hours) was used as dilution water.In the iron experiment, we used naturally enriched tap water after a 2-day stagnation period in the old pipes of the distribution system.Subsequently, the measurements were done in the first-draw sample (1.5 litres) of the stagnant tap water.The concentration of iron in the first-draw tap water sample was on average 0.79 mg/l (0.45 to 1.12 mg/l).Copper was chosen to test the manufacturers' claims for reduction of heavy metals, and ironbecause it significantly affects the organoleptic characteristics of drinking water.-Nitrate nitrogen (NO3-N) and phosphate phosphorus (PO4-P).The retention efficiency of the anions was assessed in aged and naturally enriched with nutrients aquarium water with average concentrations as follows: NO3-N -9.5 mg/l and PO4-P -1.3 mg/l.The amount of the nutrients in the water was determined spectrophotometrically on WTW Photo Lab 7100Vis, by the Spectroquant method 14773 and 14848 of Merck.-Additionally, to determine the effect of filtration/ sequential filtration on pH and the electrical conductivity of the water, their values were determined before and after each filtration.A pre calibrated pH meter Combo check HI98129 (Hanna Instruments) was used for simultaneous reading of pH, electrical conductivity and water temperature.The efficiency of water purification from pollutants (α) was determined by the formula: where: Csthe concentration of a given pollutant in the source water; Cthe concentration of the pollutant in the filtrate.

Results and discussions
The water samples were analysed for selected water quality parameters before and after passing through the water filters.The effect of water filter age and contact times were evaluated separately based on the parameters of electrical conductivity and pH, concentration of metal ions, and anionsnitrates and phosphates.To assess the effect of filter aging on the physicochemical indicators of water and the efficiency of removal of pollutants (purification efficiency), we have used filters at three distinct agesnew filters (0% lifespan); filters at 50% lifespan (100 litres) and exhausted filtersat 100% of their projected lifespan, according to the recommended by the manufacturer exploitation time.

Influence of the pitcher filters on the electrical conductivity (EC) and pH of water.
The EC of the model waters ranged between 97 and 299 μS/cm.The differences in the EC values were due to the different compositions of the model waters, as some of them were prepared by augmentation of tap water and otherswere aged aquarium water.The EC of the filtered water was reduced to a different degree, depending on the age and contact time (Figure 1), due to the removal of a part of the conductive ions by the filter.A similar effect is reported by Skoczko & Szatyłowicz [15] who report a decrease in mineralization due to the precipitation of mineral substances in the pores of the filter media.The effect on EC of the filtrates ranged between 10% and 15% reduction for a single run through the filters.The new filters had a stronger effect on the EC of water with a subsequent reduction of efficiency with filter age.In the experiment with successive filtration of the model waters the results show increased retention of ions with each passing of the water through the filters.The lowest efficiency after the tenth filtration is observed in the test with 50% exploitation capacity of the filter -25.8% reduction of the initial EC.Although the new filter and the one at 100% exploitation capacity had more than 30% difference in the single run efficiencies, they showed approximately the same effect on the EC after ten runs of the model waters through the filters (Figure 1).The pH of the model waters ranged between 6.4 and 7.09 (average 6.76).The slightly acidic nature of the model waters was due to the addition of copper sulphate.An increase in pH was observed after each passing of the model waters through the filters up to the 7th consecutive filtration.We observed a stronger increase in the pH values in model waters with lower initial values of EC, supposedly due to the higher buffering capacity of the model waters with higher initial EC values.Thus, in the experiment with aquarium water, which had the highest EC, the effect on pH was negligible -0.06 pH units.In the rest of the experiments, the effect was an increase by an average of 0.4 pH units, from an average of 6.76 to 7.16 (maximum to 7.35).There were no significant differences in the influence of filter age on the pH values.Much stronger effects on pH are observed by Doré et al. [16], who reported both increases and decreases in the pH values of the treated water, depending on the studied type of faucet-mounted point-of-use filters, or pitcher filter jugs.According to the authors, the different changes in pH values are due to the different filtering media.Krolag et al. [17] also report a decrease in the pH of filtrates from pitcher filters designed to decrease water hardness.Apart from the difference in pH values, they also observed a stronger decrease in conductivity (35%) from groundwater and well samples compared to our results.

Removal efficiency of copper and iron via pitcher filters
Figure 2 shows the removal efficiency of copper and iron ions by filters with different ages.The pitcher filters have very high elimination efficiency of copper ions from water, with both new and exploited at 50% of their capacity filters, averaging 98% removal after the first filtration.At the end of the exploitation period (after 200 litres of filtered water) the efficiency of copper removal was just 55%.Our results are consistent with the reported range of efficiency (65% to 99.8%) for the elimination of copper ions with pitcher filters [18].The removal efficiency of iron is much lower and even with new filters the reduction of iron from the water is ≈ 25% in a single filtration.This effect gradually decreases with filters age, so that at 50% of their capacity the reduction is 16% and for the exhausted filters (100%)just 6%.The decrease in the removal efficiency of copper and iron over time is a consequence of the depletion of the sorption and ion exchange capacity of the media in the filters [19].Although the model water in the experiment has concentrations of copper that are unlikely to be encountered in tap waters, the approximately 50% loss of efficiency of the filters by the end of their exploitation lifespan, as well as the more prominent decrease in the case of iron, shows people should avoid using exhausted filtrates to prevent potential health problems.
In the experiment with sequential filtration of the model waters, the efficiency of the new filter and the one at 50% lifespan reached 100% elimination of copper from the water after three filtrations.The lowest efficiency is observed for the exhausted filter, but still, they exceeded 95% reduction of copper ions after the 5 th filtration and reached 99.7% after the 10 th filtration.Different results were observed in the experiments with the sequential removal of iron in comparison to copper.The strongest loss of efficiency was observed in the 1 st and 2 nd filtrations -55% loss in comparison to the new filters.With sequential filtration, the old filters gain efficiency and by the 10 th filtration, the difference with the new filters is reduced to 33%.Doré et al. [16] report that all tested pitcher filters increased the iron, magnesium, sulphur, and zinc concentrations in the filtrate water, due to their presence in, and subsequent release from the filter media.The results from the sequential filtration experiment are in accordance with the previous research of Ndé-Tchoupé et al., [19] and Barkouch et al. [20], who showed that, apart from the material filling the cartridge, the efficiency of metal removal can be related to factors such as increased length of the filtration bed, i.e. the increased contact time between water and filter media.In the report of Doré et al. [16], where pitcher filters were used to remove lead from drinking water, the removal efficiency was in the range of 10.9% to 92.9%.

Removal efficiency of phosphate phosphorus and nitrate nitrogen via pitcher filters
Figure 3 shows the removal efficiency of PO4-P and NO3-N by filters with different ages.In our experiments, we observed an increase of 2.9% in the concentration of PO4-P after filtration through pre-conditioned new filters.This may be due to the release of phosphorus from the filter media in the initial period of its activation.The maximum amount of phosphorus is released by the 3 rd sequential filtration, after which it is sorbed again by the filter media, and by the 10 th filtration, a net removal of phosphates is observed.This shows that the prescribed initial soaking and two/three times rinsing/ washing out of the filter may not be enough for the preconditioning of the filters.To achieve the latter, it may be necessary to either prolong the soaking time or to increase the washing cycles of the filters.
Due to the above-mentioned peculiarities of the new filters, the highest removal efficiency during single filtration was observed in the experiments with filters at 50% of their lifespan -22%.The efficiency decreases over time, and for exhausted filters, it is just 8.2%.The efficiencies of nitrates removal after single filtration through a new filter, as well as through exhausted filters are similarapproximately 40%.For this reason, we have not run an experiment with filters at 50% of their lifespan.In the sequential filtration experiments the filter with a 50% lifespan, reached an efficiency of 42% phosphates reduction by the 10 th filtration.The efficiency of the exhausted filters (21%) by the 10 th filtration barely reaches the reduction of phosphates in the single filtration experiment with filters at 50% lifespan.In the sequential filtration experiment with nitrates, the new filter reached an efficiency of over 75% after just two filtrations, while the exhausted filter exceeded 70% efficiency only after the 7 th filtration.Contrary to our results Al-Haddad et al. [21] observed a reduction in the concentrations of major cations and anions, including NO3, only in the case of reverse osmosis, but not in the case of pitcher filters.Krolag et al. [17] report more than 80% reduction in the amount of nitrates in the treated waters after filtration with pitcher filters.Their results are close to the value we obtained after two consecutive filtrations of the model water.The authors did mention that the filters they used contain, as in our experiment, a mix of activated carbon and ion exchange resin, but judging on the results, they might have used filters with higher contact time or higher volume of the filter media.
A common problem in the drinking water supply is the occurrence of NO3 in the shallow ground waters situated above the impermeable rocky formations in regions with intense agriculture.Most vulnerable are the flat terrain areas with impaired drainage, where the applied artificial fertilizers are predominantly transferred down, eventually reaching the first groundwater layer.Our results show that the pitcher filters can be used to improve the drinking water quality in villages with increased concentrations of nitrates in the supplied drinking water, or in the cases of drinking water supply from home wells.

Removal efficiency of suspended solids (SS) via pitcher filters
Figure 4 shows the removal efficiency of SS during filtration through filters with different ages.A laboratory culture of M. contortum was used as model water in the experiment.After single filtration of the model water, the new filters had the highest removal efficiency, reducing the amount of the SS by ≈25%.The efficiency gradually decreases, and for filters with a 50% lifespan it is ≈20%, and for the exhausted filters it is just 6.5%.This corresponds to an overall reduction in efficiency of about 19% by the end of the prescribed lifespan of the filters.
In the experiment with successive filtration of the waters, the efficiency of the filters differed significantly and abnormally from the single filtration effects on the SS values in the water.Strong fluctuations in the retention efficiencies of SS were observed, even for successive filtrations through a single filter.As we have used an algae culture of M. contortum as suspended sediments in our experiment, and because the mesh screens at the top and bottom of the filters are with an eye-opening of 80 μm it is not surprising that the efficiency is variable and inconsistent.Factors such as different sizes of algae, packing density, and loading with suspended solids from previous filtrations affect the filter's suspended solids retention efficiency.Thus, the retention efficiency of the new filter was just above 25% and increased until the 5 th sequential filtration, with a consequent sharp decline in the efficiency of Chl-a retention which by the 10 th filtration fell below 10%.Similar results were observed with the exhausted filters (100%), with a decrease to approximately 10% by the end of the sequential filtration.The 10% retention of the algae seems to correspond to the size fraction of the population of M. contortum which are permanently retained by the filters, while the smaller cells pass through the filters with some being retained by the fibrous material in the filter.Only in the experiment with the filters at 50% lifespan, we observed a steady increase in the retention efficiency after the second filtration.This may be due to the different packing densities of the filter media.Al-Haddad et al. [21] report that measurable values of turbidity and TSS were observed in the outlet samples of all types of water filters except for the filter cartridge with 5 μm mesh screens.This shows that the efficiency of the pitcher filters to retain suspended solids depends on the mesh screen size and not so on the filter media.

Conclusions
This study demonstrates that the use of pitcher filters results in a notable decrease in conductivity values and concentrations of various pollutants, more specifically, the concentrations of metals such as copper decreased by 98%, and iron by 25%.Additionally, major anions exhibit reductions, with NO3-N decreasing by 40% and PO4-P by 9%.The application of pitcher filters also shows a moderate impact on suspended solids, with a 17% decrease in SS.Furthermore, the filtration process contributes to the stabilization and elevation of pH values, and this effect is inversely proportional to the water's buffering capacity.The removal efficiencies/observed effects increase with prolonged contact time between water and filter media, while they decrease with the aging of the filters.The decline in removal efficiency with filter age is lowest for nitrates, while for the other investigated indicators, it ranges from 30% to 60%, especially in the second half of the filters' exploitation lifespan.In summary, the utilization of pitcher filters proves to be an effective method for enhancing water purification at home.These filters have the potential to regulate pollutants, notably metals and nitrates, which pose particular challenges in regions relying on water from underground sources.

Figure 1 .
Figure 1.Effect of filtration (percent reduction) on electrical conductivity and pH of water.Left panel -relationship between the effect on EC and number of sequential filtrations.Right panel -average values of water pH after different number of sequential filtrations.

Figure 2 .
Figure 2. Removal efficiency of metals: left panelremoval efficiency of copper with filters with different age and after sequential filtrations; right panelaverage efficiency of iron removal with standard deviations after sequential filtrations.

Figure 3 .
Figure 3. Removal efficiency of dissolved ions: left panelremoval efficiency of phosphates with filters with different age and after sequential filtrations; right panelaverage efficiency of nitrates removal with filters with different age and after sequential filtrations.

Figure 4 .
Figure 4. Removal efficiency of suspended solids with filters with different age and after sequential filtrations.