Treatment of waste stabilization pond effluent using natural zeolite for irrigation potential

Direct utilization of treated effluent from natural treatment systems for irrigation can be challenging on sensitive plants due to high levels of salinity. Post-treatment of such an effluent prior to its applicability in irrigation can be of significant importance. In this study, the wastewater from a natural treatment plant was treated using a lab-scale filtration system with zeolite as a filter material. Three different column depths (0.5 m, 0.75 m, and 1 m) were used to investigate the effect of column depth on the treatment efficiency of the media. The suitability of the raw wastewater and the treated effluent from each column for irrigation purposes was investigated. The water quality parameters investigated were; electrical conductivity (EC), total dissolved solids (TDS), sodium (Na+), calcium (Ca2+), and magnesium (Mg2+). From the analysis results, it was observed that the column depth had a significant influence on the removal efficiency of the pollutants. The highest removal efficiency (94.58%) was achieved from the combination of electrical conductivity and 1 m column depth, while the lowest removal efficiency (10.05%) was observed from the combination of calcium and 0.5 m column depth. The raw wastewater fell mostly into a “very high” hazard, which is class four (C4) based on electrical conductivity and class four (S4) based sodium adsorption ratio; making it unsuitable for irrigation purposes. However, when the wastewater was subjected to 1 m column depth, the quality of the treated effluent improved significantly which in turn also improved the suitability of the effluent for irrigation purposes, with percent compliance ranging from 20.19% to 97.54%.


Introduction
adsorbents is regarded as among the relatively economical solutions for treating and reusing wastewater of high salinity [23]. Zeolites as a filter material provide one of the economical technologies and have been widely used in the field of water treatment as ion exchangers and adsorbents [24,25].
In this study, the potential applicability of zeolites on treating the effluent from a waste stabilization pond for irrigation purposes of low salinity tolerance plants is investigated. Three different column depths (0.5 m, 0.75 m, and 1 m) are used the investigate the influence of column depth on the treatment efficiency of zeolite. Then the effluent from each column is investigated for its potential applicability in irrigation, especially for sensitive plants.

Case study description, analytical methods, and wastewater characteristics
The wastewater samples used in this study were collected from the Vingunguti wastewater stabilization ponds in Dar es Salaam, Tanzania, approximately 7.2 km from the city Centre (latitude: 6˚50'17.20"S, longitude: 39˚14'8.62"E), under the permit given by the Dar es Salaam Water and Sewerage Authority (DAWASA).
Several water quality parameters were investigated in this study, including; EC, Na + , TDS, Mg 2+ , and Ca 2+ . The selection of the parameters is based on their significance in determining water suitability for irrigation purposes. In general, Na + was measured using the Sodium-Ion Selective Electrode Method [26], while both Mg 2+ and Ca 2+ in water samples were measured using the Ethylenediaminetetraacetic Acid (EDTA) Method [27], with Na 2 EDTA 0.05, Acetylacetone, Tris(hydroxymethyl)-aminomethane (TRIS), Distilled Water and Electrolyte solution L300 as reagents. TDS and EC were determined using the TDS Meter Digital Water Tester (Lxuemlu, Shenzhen, China). Table 1 presents a summary of the raw wastewater characteristics in terms of minimum (Min), maximum (Max) median, arithmetic mean (AM), and standard deviation (STD) as well as the guidelines for irrigation water as recommended by the Food and Agriculture Organization of the United Nations (FAO) [28]. A maximum EC concentration of 4224 μS/cm was observed from the raw wastewater, with an average concentration of 2478.1 μS/cm.

Experimental setup
Three different fixed-bed columns with 0.5 m, 0.75 m, and 1 m depths were used to investigate the influence of column depth in the treatment efficiency of natural zeolite (Fig 1). The selected column depths serve as both lab-scale and small-scale wastewater treatment systems. The column containers are of Polyvinyl chloride (PVC) material with approximately 5.08 cm in diameter (Table 2). All three columns were packed with natural zeolite adsorbents (clinoptilolite) composed of a microporous arrangement of silica and alumina tetrahedra with an average particle size of 1.5 mm (FM Stock and Supplies, Kenmare, Gauteng, South Africa).

PLOS ONE
To allow equal distribution of flow in the columns, the top surfaces of the columns were covered by perforated plates with evenly distributed holes. 100 L storage drum was used to feed the columns at a controlled rate of 0.0035 L/s (based on the amount of wastewater, size of the filter columns, and anticipated daily working hours). To maintain all the solids in suspension, the wastewater was slowly and continuously stirred. The Wet-packing approach of the porous medium was used with the purpose of minimizing layering and air entrapment inside the filing. All three columns were mounted vertically, and glass wool was used at the bottom of the column acting as supporting material of the adsorbent bed. After packing the column, deionized water was passed through the column for some time, followed by the introduction of the feed water. The filtrate samples were collected at a regular time interval. All the experiments were carried out at room temperature (20 to 25 0 C). Table 2 provides a summary of the physicochemical characteristics of the zeolite filter materials used in the study.

Relationships among parameters.
Correlation analysis was among the approaches used to analyze results from the experiments. Correlation matrices were developed to evaluate the strength of the relationship among the studied parameters. A high correlation indicates that two or more variables have a strong relationship with each other. While a weak correlation indicates that the variables are hardly related. Table 3 provides a summary of the interpretation of the correlation indices used in this study.

Data distribution analysis.
Apart from the correlation matrices, box and whisker plots were used to evaluate data distributions among the water quality parameters. The evaluation is based on the distribution of numerical data and skewness through data quartiles (percentiles) and averages. In general, box plots show the five-number summary of a set of data: the minimum score, first (lower) quartile, median, third (upper) quartile, and maximum score. In this study; EC, Na, Mg, Ca, and TDS were analyzed using box and whisker plots.

Adsorption capacity analysis.
The adsorption capacity of the filter material (q) was investigated based on the ratio of the amount of adsorbate adsorbed in mg and the amount of adsorbent used for adsorption expressed in gm. Eq 1 provides a summary of the formula used in the computation of the adsorption capacity for the investigated water quality parameters. Some other parameters such as the weight of the material used (M) and the volume (V) of the wastewater subjected to the treatment system were also used.
Where; C r is the average concentration in the raw wastewater and C t is the average concentration in the treated effluent.

Salinity hazard analysis.
The salinity hazard zones based on electrical conductivity (EC) were classified into four classes; class one (C1) class two (C2), class three (C3), and class

PLOS ONE
four (C4), ranging from low salinity (C1) to very high salinity (C4). Table 4 provides a summary of the salinity hazard zones based on EC with their interpretations in terms of usability [29]. Table 5 provides a summary of the salinity hazard zones based on Sodium adsorption ratio (SAR) with their interpretations in terms of usability. The salinity hazard zones based on SAR were also classified into four classes and used in the Wilcox diagrams; class one (S1) class two (S2), class three (S3), and class four (S4) ranging from low sodium hazard (S1) to very high sodium hazard (S4) [30].
SAR can be defined as an index used to define the effect of sodium concentration in a sample in relation to calcium and magnesium. More specifically, the SAR index is achieved by diving the square root of 1/2 of the calcium plus magnesium concentrations. Eq 2 provides a summary of the formula used in the computations of SAR [31].

SAR ¼
Na þ ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi Furthermore, Wilcox diagrams were plotted from the raw wastewater, effluent from 0.5 m, 0.75 m, and 1 m zeolite columns. The Wilcox plot is a semi-log scatter plot of the "sodium hazard" (SAR) on the Y-axis versus the "salinity hazard" (EC) on the X-axis. It must be noted that the EC is plotted by default on a log scale. The treated effluent suitability for irrigation mainly depends on the concentration of total salinity and sodium related to other ions [32]. Therefore, the diagrams were used to evaluate the risk levels in the raw wastewater and the treated effluent from the three columns.
The irrigation water quality standards were selected to investigate further the quality of the treated effluent. Eq 3 gives a summary of the approach used for the percent compliance calculations.
where; C p , percent compliance, S i , the recommended standard for an i th parameter, C i , the concentration of the i th parameter.

Water quality characterization
The analysis of the water samples before and after the treatment was successfully executed. In the raw wastewater samples, the average EC concentration was 2478.1 μS/cm, while that of TDS was 1622.20 mg/L. The EC concentration in the raw wastewater falls in class four (C4) based on the salinity hazard zones, with an indication that the effluent from waste stabilization ponds cannot be used directly for irrigation purposes. Average concentrations of 131.9 mg/L, 14.6 mg/L, and 9.3 mg/L were recorded from Na + , Mg 2+ , and Ca 2+ , respectively. From Fig 2, it can be observed that from the EC boxplot, the median line is closer to the middle, indicating that the EC data distribution is symmetric or normal. From the Na + boxplot, it can be observed that the median line closer to the upper quartile with an indication that the distribution of Na + data in the raw wastewater is "negatively skewed". This means sodium data constituted a higher frequency of low concentration values than the high concentration values.
As observed from the Na + boxplot, a similar case applies to Mg 2+ concentration data distribution from the raw wastewater. While, from the Ca 2+ boxplot, the median line is observed to be closer to the lower quartile meaning that the water quality data constitute a higher frequency of more high concentration values than the low concentration values also known as "positive skewness". Similarly, from the TDS boxplot, the median line is observed to be closer to the lower quartile meaning that the water quality data constitute a higher frequency of high concentration values than the low concentration values ("positively skewed").
A correlation matrix for the studied water quality parameters in raw wastewater was developed to evaluate the strength of the relationship among them. From Table 6, it can be observed that the general correlation among the parameters ranges from a "strong" to a "very strong" relationship. The highest correlation index of 0.966002 was achieved between Na and EC, followed by 0.945631 between TDS and Na. Also, a very strong correlation can be observed between TDS and EC with a correlation index of 0.944682. The lowest correlation index can be observed between the Mg 2+ and Na + with a correlation index of 0.598734. However, the index between Mg 2+ and Na + falls under a strong relationship.
In the literature, other studies have also observed a very strong relationship between TDS and EC to the point of recommending the EC to estimate TDS based on the linear relationship as shown in Eq 4 [33,34]. With the fact that TDS measurement is considered to be a time-consuming process, simplicity is often estimated from EC assuming TDS are predominantly ionic species of low enough concentration to produce a linear TDS-EC relationship [35].
Where; K e is a proportionality constant ranging from 0.54 to 1.1. According to Thirumalini and Joseph [36], that investigated the correlation between EC and TDS in natural waters, it was observed that the correlation index between TDS and EC was 0.63 for samples taken from pollution-free residential areas as well as ranging from 0.59 to 0.93 for samples taken from the textile industrial belt. Therefore, the general strong correlation between TDS and EC observed in the literature agrees with the results obtained from this study.
From Tables 7-9, it can be observed that, when the depth was increased from 0.75 m to 1 m, there was a slight difference in terms of EC and TDS removal in the wastewater. The average EC concentration from the 0.75 m column depth was 487.85 μS/cm, while from the 1 m column depth, the average EC concentration was 378.51 μS/cm. Also, the average TDS

PLOS ONE
concentration from the 0.75 m column depth was 134.36 μS/cm, while that of 1 m column depth was 130.163 μS/cm. The phenomenon suggests that the treatment approach has a gradual removal efficiency as the column depth increases from 0.75 m to 1 m.
From Fig 3, the EC, Na + and TDS boxplots, the median line closer to the upper quartile with an indication that the distribution of EC, Na +, and TDS data in the treated effluent using the 1 m column of zeolite is "negatively skewed". While that of Mg 2+ is observed to be closer to the middle, indicating that the Mg 2+ data distribution was symmetric or normal. From the Ca 2 + boxplot, the median line is seen to be closer to the upper quartile with an indication that the distribution of Ca 2+ data in the treated effluent using the 1 m column of zeolite is "negatively skewed". This means the Ca 2+ data constituted a higher frequency of low concentration values than the high concentration values.

Removal efficiencies
The removal efficiency of the studied parameters was observed to be significantly affected by the column depths; the more the column depth, the higher the removal efficiency (Fig 4). The highest removal efficiency (94.58%) was achieved from the combination of EC and 1 m column depth. This was followed by a removal efficiency of 91.98% from the combination of TDS and 1 m column depth. The lowest removal efficiency can be observed from the combination of Ca 2+ and 0.5 m column depth. In the literature, natural zeolite has also been observed to be highly efficient in terms of TDS removal. According to [37], which investigated the treatability of brackish groundwater by zeolite filtration in Sumur Tua Wonocolo, Kedewan, Bojonegoro, East Java, a TDS removal efficiency of up to 84% was achieved.

Influence of pH on the general performance of the treatment systems
The influence of pH in the general performance of the filter material for the removal of dissolved solids was investigated based on the correlation indices retrieved from the computed correlation matrices between pH and the water quality parameters. Fig 5 presents the correlation indices in the range of 0 to 1, with 0 being the lowest relationship and 1 being the highest relationship; for total dissolved solids (TDS), calcium (Ca), magnesium (Mg), sodium (Na), and electrical conductivity. The correlation indices ranged from approximately 0.62 to 0.92, which can be termed as a strong to very strong range of correlation. The investigation of the potential influence of pH in the performance of the filter materials in this study is based on the monitored initial pH in raw wastewater. Also, with the fact that pH in the raw wastewater ranged from 5.6 to 8.5, the results were observed when the pH was close to neutral. The phenomenon can be highly linked to the fact pH has a tendency to promote electrostatic interactions between the ions of the pollutant to be adsorbed and the sorbent materials, a phenomenon that makes pH being a crucial parameter in controlling the adsorption processes; as also observed from the study conducted by Ruth Sánchez-Hernández et al., [38].

Filter material adsorption capacity based on the investigated water quality parameters
From Fig 6 it can be observed that there was no specific trend in terms of absorption capacity as influenced by the column depth. Probably, the phenomenon is linked to the fact that the filter material characteristics used in all three columns were similar. For instance, from TDS removal when the wastewater was subjected to the 0.5 m column depth, the average adsorption capacity was 49.96 mg/g, 57.68 mg/g when the wastewater was subjected to the 0.75 m column depth, and 57.55 mg/g when the wastewater was subjected to the 1 m column depth; whereby, on average, 55.06 mg/g adsorption capacity was achieved. From sodium removal, 9.91 mg/g of adsorption capacity was achieved when the wastewater was subjected to the 0.5 m column depth, 2.63 mg/g when the wastewater was subjected to the 0.75 m column depth, and 4.31 mg/g when the wastewater was subjected to the 1 m column depth. While, from magnesium removal, 2.23 mg/g was achieved from 0.5 m column depth, 3.14mg/g from 0.75 m column depth, and 4.69 mg/g from 1 m column depth. Moreover, from calcium removal, the adsorption capacity of the filter material ranged from 1.63 mg/g to 3.46 mg/g.
The breakthrough curves for all the investigated column depths and water quality parameters were also plotted based on the ratio of the concentration in the treated effluent at a specific time and the initial concentration in the raw wastewater versus time. From Fig 7, it can be observed that there is a potential relationship between the column depth and the breakthrough times; whereby, when the column depths increase, the breakthrough times also increase. The

PLOS ONE
phenomenon can be likely linked to the fact that the increase in column depth provides space for more filter materials that also take longer to reach the breakthrough time.

Percent compliance
Percent compliance analysis was also executed as part of investigating the quality of the treated effluent with respect to the irrigation water quality standards set by FAO. From Table 10, it can be seen that in the raw wastewater the concentrations of EC, TDS, and Na were above the recommended limits with -65.21%, -35.18%, and 1365.89% compliance, respectively. However, when the wastewater was subjected to 0.5 m column depth, EC and TDS were observed to be complying with the standards having 9.81% and 7.73% compliance, respectively except for Na. Also, when the wastewater was subjected to 0.75 m column depth, the percent of EC and TDS increased to 67.48% and 89.15% respectively, whereby, the percent of Na + increased from -1082% when the wastewater was subjected to 0.5 m column depth to -130.56% from 0.75 m column depth. However, when the wastewater was treated using the 1 m column depth all the investigated water quality parameters were within the recommended limits with percent compliance ranging from 20.19% to 97.54%.

Irrigation potential analysis
From Fig 8, it can be observed that the raw wastewater falls under high (C3) to very high (C4) hazards based on the EC. Approximately 68% of the values fall under the very high hazard while 32% fall under the high hazard category. While, based on the SAR, the raw wastewater falls from low (S1) to very high (S4), with most of the values falling under medium (S2) and high (S3). The general phenomenon suggests that raw wastewater is not recommended for irrigation purposes especially for low-salt-tolerance plants based on EC and unsatisfactory for most of the crops based on SAR. From Fig 9, it can be observed that the treated effluent from 0.5 m column of zeolite falls under high (C3) to very high (C4) hazard based on the EC. Approximately 97% of the values fall under the high hazard while 3% fall under the very high hazard category. While, based on the SAR, the treated effluent from 0.5 m zeolite column falls under low (S1) to high (S4), with most of the values falling within medium (S2) hazard. The general phenomenon suggests that the treated effluent from the 0.5 m column is under appreciable hazard but can be used with appropriate management based on SAR and can be used for irrigation purposes with some management practices based on EC.
From Fig 10, it can be observed that the treated effluent from the 0.75 m zeolite column falls under medium (C2) hazard based on the EC. Almost 100% of the values fall under medium hazard. While, based on the SAR, the treated effluent from 0.75 m falls under low (S1), with very few values getting close to medium (S2). The general phenomenon suggests

PLOS ONE
that the treated effluent from a 0.75 m column can be used for crop irrigation purposes with little or no hazard based on SAR and can be used with moderate leaching based on EC.
From Fig 11, it can be observed that the treated effluent from the 1 m zeolite column falls under a low (C1) hazard based on the EC. While, based on the SAR, the treated effluent from 1 m column under low (S1), with very little values getting close to medium (S2). The general phenomenon suggests that the treated effluent from the 1 m column can be used for crop irrigation purposes with little or no hazard based on SAR and can be used safely based on EC.

Conclusions
In this study, the potential applicability of zeolites on treating the effluent from a waste stabilization pond for irrigation purposes has been investigated with three different column depths. A correlation among the studied parameters was observed with the highest correlation index of 0.966002 achieved between Na + and EC, followed by 0.945631 between TDS and Na + . Also,  the results showed that the pollutants removal efficiency increased with the increase in column depth. Among the studied parameters, the highest removal efficiency (94.58%) was achieved from the combination of EC and 1 m column depth, while the lowest removal efficiency (10.05%) was observed from the combination of Ca 2+ and 0.5 m column depth. From the hazard analysis, the raw wastewater generally fell into the "very high" hazard class based on both EC and SAR. In that matter, the raw wastewater must be treated further prior to its application for irrigation purposes. However, the status improved after the treatment using different column depths. In that matter, the results from this study revealed further that, it is always important to investigate the quality of effluent from natural treatment systems before subjecting it to any sort of irrigation. Moreover, the use of zeolite can provide one of the efficient approaches to improve the effluent and make it suitable for irrigation.