Abstract

The treatment works under investigation uses a recovered wash water plant (RWWP) to remove impurities prior to recycling filter backwash water. Filter backwash water (raw water) is characterized by high quantities of cyanobacteria cells and associated organic compounds; a potential threat when recovered water is recycled. The aim of this study is to identify the cyanobacteria cells and associated organic compounds in the filter backwash water and to subsequently evaluate the effectiveness of the RWWP in removing these organic impurities during the following periods; autumn-winter and spring-summer. Results showed that at least six major phytoplankton groups were present in the filter backwash water with turbidity levels (59 and 46 NTU; autumn–winter and spring-summer, respectively) being much higher than the drinking water productions standard of ≤5 NTU. Cyanobacteria were a dominant group (mean of 80% and above) in the total phytoplankton composition of the raw water and consisted of three genera (Anabaena sp., Microcystis sp. and Oscillatoria sp.), which were effectively removed by the RWWP (up to 99%). However, associated organic compounds such as geosmin, total organic carbon (TOC), dissolved organic carbon (DOC) and microcystin were not effectively removed during the different seasonal periods but were of such low concentrations that they posed no major risk to the drinking water quality, meeting the RWWP water quality standard.

HIGHLIGHTS

  • The seasonal occurrence of algae and cyanobacteria in recycled backwash water and the effective removal thereof.

  • The correlation between microcystin and Microcystis sp. in recycled filter backwash water with chlorophyll pigments removal confirming the removal efficiencies of algae and cyanobacteria.

  • Poor removal efficiencies of organic compounds by a recycled water treatment plant require optimization practices.

  • Implement phytoplankton water quality monitoring for RWWP and the implementation of a routine sampling plan to determine water quality at different stages of the RWWP.

  • Stringent water quality monitoring plans to be implemented during all seasonal periods in the RWWP, especially where organic compounds associated with cyanobacteria are a concern.

INTRODUCTION

Water is a limited and critical resource to achieve sustainable goals that include economic, environmental and social development (DWAF 2002). Access to safe drinking water is an essential component of effective policy for health protection and sanitation, as well as for quality of life (Matthews & Bernard 2015). It is evident from international literature that many countries are facing challenges in securing water for basic human needs and development (WEF 2017). According to Grey & Sadoff (2007), achieving water security by reducing its destructive potential and increasing the productive potential will remain a central challenge for poor countries of the world today; however, it has always been a goal of human society to be achieved. Furthermore, the degradation of water quality is putting more pressure on water security, especially in water-scarce countries such as South Africa. Good health and well-being may not be possible without access to adequate safe drinking water that meets the needs of an entire population (Govender 2016).

Population growth and several other factors such as mining, urbanization, industrialization, agriculture and power generation negatively impact on water quality. This leads to cultural eutrophication where source water becomes high in nutrient concentrations, such as nitrogen and phosphorus, supporting high phytoplankton growth rates and biomass (Oberholster 2010; Swanepoel et al. 2017). Cyanobacterial blooms are a threat to the supply of drinking water in South Africa, especially in areas that have shortages of water treatment facilities or where infrastructure is not functioning properly (Van Ginkel 2011). Cyanobacteria, also known as blue-green algae, produce a range of cyanotoxins (Van Ginkel 2011) such as hepatic-, neuro-, and cytotoxins, which are mostly known from the cyanobacteria genera Microcystis sp., Oscillatoria sp., Aphanizomenon sp., Anabaena sp., Nodularia sp., Nostoc sp. (Harding & Paxton 2001) and Cylindrospemopsis sp. (Van Ginkel 2011). The cyanotoxins produced by cyanobacteria have an effect on human health, such as poisoning incidents that might be acute or sub-acute liver damage caused by the largest group of toxins known as the microcystins (Harding & Paxton 2001). Cyanobacteria produce taste and odour compounds, the most common of which are geosmin and 2-methyl-isoborneol (2–MIB). These organic compounds may cause degradation of aesthetic quality of water yet, even though it is unpleasant, they are not harmful but can lead to consumer complaints (Swanepoel et al. 2008).

The rapid growth of algae in response to elevated concentrations of nutrients in water bodies may clog filters and block screens (Ewerts et al. 2013). Some cyanobacteria, such as Microcystis sp. can produce coagulant inhibitor proteins that may disturb the coagulation process (Chorus & Bartram 1999). Increasing the cost of treating water (Sibande 2013; Chinyama et al. 2016) and potentially penetrating the final treated water (Swanepoel 2015), giving water a greenish colour due to regrowth in the distribution system (DWAF 1996). According to Hitzfeld et al. (2000), coagulation can be an efficient method for eliminating cyanobacterial cells from water, whereas soluble cyanotoxins are not very efficiently removed by this method. Conventional water treatment plant processes (flocculation, sedimentation, filtration and chlorination) might not be effective in removing dissolved substances such as microcystins derived from Microcystis sp. (Oberholster et al. 2004; Ewerts et al. 2013; Swanepoel et al. 2017). In addition, when cells are present in treated water during disinfection, they may serve as precursors for the formation of harmful trihalomethanes. Trihalomethanes form when organic material reacts with chlorine during pre-oxidation (DWAF 1996).

It is important for water treatment facilities to be able to manage and monitor cyanobacteria and related organic compounds in the drinking water treatment processes. The presence of cyanobacteria affects the water treatment and the quality of water in general. South Africa is a country with water shortages (DWAF 2002), making it important to recover filter backwash water, since filter backwashing uses large volumes of treated water. Filter backwashing is an integral part of the treatment process and plant operation to clean filters periodically using treated water combined with air in a reverse direction to the normal flow. This produces a significant amount of contaminated backwash water containing high concentrations of total suspended solids (TSS), total organic carbon (TOC), inorganic precipitates as well as microorganisms such as phytoplankton cells (Li et al. 2017). It is therefore important to treat the recovered water to a good quality to allow for distribution to consumers for drinking. Poorly treated or untreated recovered water may have a negative impact on the final treated water; therefore, the following considerations should be given; (1) contaminants in recovered water could be more efficiently removed than those in raw water because the particles in recovered water have already been destabilized, and (2) recovered water increases the collision and adhesion probabilities for the suspended particles in the coagulation process thus further enhancing the subsequent treatment processes.

This study is aimed at understanding and gathering information regarding the efficacy of a recovered wash water plant (RWWP) associated with the main water treatment plant (main WTP) in removing cyanobacteria and associated organic compounds from the recovered water. The objectives of the study were to (1) identify the phytoplankton groups in the recovered filter water supplied to the RWWP, (2) identify cyanobacteria genera present in the recovered filter water, and (3) investigate the effectiveness of the RWWP in removing cyanobacteria and associated organic compounds during autumn-winter and spring-summer sampling occasions.

METHODS

Study area

The study was conducted at the 35 ML/D RWWP outside of Vereeniging (Gauteng Province, South Africa). The plant applies conventional water treatment processes that include coagulation, flocculation, sedimentation and sand filtration to treat recovered filter backwash water. After treatment, recycled water is mixed with water in the main water treatment plant for disinfection and distribution.

Coagulant chemicals used at RWWP

The following coagulant chemicals and dosages were dosed in the RWWP during the study period: polyelectrolyte 3,835 was dosed as the primary coagulant and 45% FeCl3 as a secondary coagulant. The optimum dosages, as determined by jar test experiments, were 5 mg/L polyelectrolyte and 4 mg/L FeCl3.

Sampling and water quality analysis

Raw water samples: for the purpose of this study, raw water refers to recovered filter backwash water destined for treatment by the RWWP. Raw water samples were collected as the first samples prior to entering the RWWP. Subsequent process samples were collected after sedimentation and then after filtration. All raw water and subsequent process samples were collected during March 2017–February 2018, referring to the sampling occasions as autumn-winter and spring–summer. The analysis of water samples was performed at an accredited analytical laboratory situated in Vereeniging, Gauteng, South Africa. The laboratory is accredited by SANAS (South African National Accreditation System).

Phytoplankton analyses (identification and enumeration) was determined by sedimentation/centrifugation technique that was originally described by Lund et al. (1958) and adapted by Swanepoel et al. (2008). Chlorophyll (Chl-665) was determined by spectrophotometry according to the Swanepoel et al. (2008) method. The taste and odour compound geosmin was extracted using solid phase extraction and quantified by means of the gas chromatography-mass spectrometry (GCMS) method (APHA 2013) and adapted for the laboratory according to Swanepoel et al. (2008). The concentration of the microcystin (total microcystin) was determined by means of the enzyme linked immune sorbent assay (ELISA) technique, also described by Swanepoel et al. (2008), using Envirologix ELISA kits (Envirologix 2013). The content of organic matter was determined as total organic carbon (TOC) and dissolved organic carbon (DOC) using standard methods for examination of water and wastewater samples (APHA 2013). TOC samples were collected headspace-free in 40 mL glass vials and preserved with 21% phosphoric acid and measurements performed (Phoenix and Fusion). The physical parameters turbidity, pH and electrical conductivity were measured using a HACH-2100AN turbidimeter, Metrohm 691 pH meter and LF 538 conductivity meter, respectively.

Percentage removal during raw water (RWWP) and after filtration for the study period was determined by the following formula: 
formula

Flow of recovered filter water to RWWP treatment processes and to main treatment works

The main WTP treat water from Vaal Dam, using conventional treatment processes (coagulation, flocculation, sedimentation, carbonation, sand filtration, chlorination and distribution) to treat approximately 1,800 ML/D. The plant uses chemicals such as hydrated lime, polyelectrolytes as coagulants and carbon dioxide (to adjust pH). After the process of coagulation, flocculation and sedimentation, the water is channelled to 3 filter houses with a total of 96 rapid gravity sand filters, which run on 36-hour backwash cycles during normal operations. Recovered filter backwash water (raw water) is pumped to the RWWP as filter backwash effluent from the main water treatment plant. After conventional treatment in the RWWP (coagulation, flocculation and sedimentation), recycled or treated water is channelled to the main WTP treated sump (Figure 1). The RWWP contains a filter house with three rapid gravity sand filters set on an automatic backwash mode. An illustration of the treatment process is provided in the diagram below (Figure 1).

Figure 1

Schematic of the RWWP showing the sequence of purification processes (coagulation, flocculation, sedimentation and sand filtration) and included sampling points, linked to the main water treatment plant.

Figure 1

Schematic of the RWWP showing the sequence of purification processes (coagulation, flocculation, sedimentation and sand filtration) and included sampling points, linked to the main water treatment plant.

Statistical analysis

Three sets of data were used for the analyses, namely raw water, after sedimentation and after filtration, and they were clustered as autumn-winter and spring-summer. Different variables were examined such as cyanobacteria types and concentration present, organic compounds such as geosmin and microcystin. The datasets were used to determine the efficiency of the RWWP in removing cyanobacteria and associated organic compounds. A single factor analysis of variance (ANOVA) was used to determine removal efficiencies at different stages of water treatment by examining for statistically significant differences. The level of significance for all statistical analysis was set at a p value ≤0.05 and the hypothesis was that there is a significant improvement in water quality.

Datasets obtained from 47 sampling occasions were used to perform a Canonical Correspondence Analysis (CCA) using the computer software package CANOCO version 4.5. The CCA was carried out to determine the relationship between parameters such as cyanobacteria (Microcystis sp., Oscillatoria sp. and Anabaena sp.) with Chlorophyll-665, microcystin, geosmin, total organic carbon (TOC) and dissolved organic carbon (DOC). Ordinations were interpreted using the following rationale: parameters are positively correlated with each other if their arrows subtend a small angle. Negatively correlated if their arrows are in the opposite directions (180°) and correlated when their arrows are at 90° angle.

RESULTS AND DISCUSSION

Assessment of raw water entering the RWWP

Six major phytoplankton groups were detected in the raw water samples during all months of the study period. The phytoplankton composition consisted of the following major groups; namely, Bacillariophyceae (diatoms), Chlorophyceae (green algae), Cyanophyceae (cyanobacteria or blue-green algae), Dinophyceae (dinoflagellates), Chrysophyceae (golden-brown algae) and Euglenophyceae (euglenoids). The total phytoplankton composition in raw water was dominated (mean of 80% and above) by cyanobacteria, consisting of the following species: Anabaena sp., Microcystis sp. and Oscillatoria sp. Microcystis sp. occurred more frequently and it was the most dominant (more than 80% of the dataset) than other species of the same group in the raw water during both autumn-winter and spring-summer months. According to Van Vuuren et al. (2006), Microcystis sp. usually dominates a habitat and excludes almost all other species of cyanobacteria, which appeared to be the case in this study. Table 1 below shows the cell concentrations of phytoplankton groups, cyanobacteria species and organic compounds as well as the concentrations of three important physical parameters measured in raw water samples during the study period.

Table 1

Average concentrations and standard deviation (±SD) for phytoplankton groups, cyanobacteria genera, organic compounds and physical parameters measured in raw water entering the RWWP during the study period

Autumn-winter
Spring-summer
Phytoplankton groupsUnitsMean± SDMean± SD
Bacillariophyceae cells/mL 769 548 909 890 
Chlorophyceae cells/mL 200 361 315 611 
Cryptophyceae cells/mL 291 241 227 130 
Dinophyceae cells/mL 52 33 40 
Euglenophyceae cells/mL 45 13 40 
Cyanophyceae cells/mL 6,593 7,602 22,653 26,244 
Cyanophyceae genera 
Anabaena sp. cells/mL 199 224 4,358 4,677 
Microcystis sp. cells/mL 6,103 7,595 18,310 24,677 
Oscillatoria sp. cells/mL 291 710 40 
Organic compounds 
Geosmin ng/L 4.0 59 66 
Microcystin μg/L 0.85 0.7 0.8 
Chlorophyll − 655 μg/L 18 17 34 20 
TOC mg/L 0.7 1.4 
DOC mg/L 1.1 0.9 
Physical parameters 
Turbidity NTU 59 20 46 13.0 
pH pH units 8.7 0.11 8.7 0.2 
Electrical conductivity mS/m 18 1.0 17.3 1.0 
Autumn-winter
Spring-summer
Phytoplankton groupsUnitsMean± SDMean± SD
Bacillariophyceae cells/mL 769 548 909 890 
Chlorophyceae cells/mL 200 361 315 611 
Cryptophyceae cells/mL 291 241 227 130 
Dinophyceae cells/mL 52 33 40 
Euglenophyceae cells/mL 45 13 40 
Cyanophyceae cells/mL 6,593 7,602 22,653 26,244 
Cyanophyceae genera 
Anabaena sp. cells/mL 199 224 4,358 4,677 
Microcystis sp. cells/mL 6,103 7,595 18,310 24,677 
Oscillatoria sp. cells/mL 291 710 40 
Organic compounds 
Geosmin ng/L 4.0 59 66 
Microcystin μg/L 0.85 0.7 0.8 
Chlorophyll − 655 μg/L 18 17 34 20 
TOC mg/L 0.7 1.4 
DOC mg/L 1.1 0.9 
Physical parameters 
Turbidity NTU 59 20 46 13.0 
pH pH units 8.7 0.11 8.7 0.2 
Electrical conductivity mS/m 18 1.0 17.3 1.0 

The average (Tables 1 and 3) diatom concentrations recorded during autumn-winter and spring–summer months showed important contributions to the total phytoplankton composition (>500 cells/mL), while other phytoplankton groups (green algae, golden-brown algae, dinoflagellates and euglenoids) were detected in relatively low concentrations (<500 cells/mL). Groups occurring in concentrations below 500 cells/mL may not pose a risk to the water quality and treatment processes. However, it remains a priority for treatment facilities, especially for drinking water treatment plants, to remove algae that may introduce any quality or treatment problems to acceptable levels at each stage of treatment. Although diatoms were present in lower quantities compared to the cyanobacteria, their removal efficacy should be monitored simultaneously with the dominant cyanobacteria group.

The average (Tables 1 and 3) cyanobacteria concentration in the raw water was recorded at 6,593 cells/mL during the autumn-winter, while an average (Tables 1 and 3) concentration of 22,653 cells/mL was recorded for spring-summer months. During spring-summer, SD values for Dinophyceae, Euglenophyceae and Oscillatoria sp. were recorded as zero, indicating that all the samples had the same values, while exceeding mean SD values were recorded for Chlorophyceae, Cyanobacteria and geosmin, indicating that the data was more spread out (Table 1). High cell concentrations are likely to affect the drinking water process when cyanobacteria enter the water treatment plant (Swanepoel et al. 2008). Thus, coagulant optimisation strategies should form an integral part of the water treatment when raw water is characterised by relatively high cyanobacteria concentrations (>2,000 cells/mL).

The average turbidity levels determined during both occasions were recorded as 59 NTU and 46 NTU for autumn-winter and spring-summer months, respectively. Turbidity is an important parameter measured during water treatment (≤5 NTU after sedimentation) and an aesthetic criterion set by the South African National Standard (SANS 241 2011). Ewerts et al. (2015) found that phytoplankton content in water may have an impact on the turbidity; however, the removal of turbidity cannot necessarily be used as an effective indicator of phytoplankton removal. Other important physical parameters, such as pH and conductivity, were measured and yielded the same results for both autumn-winter and spring–summer periods and posed no treatment challenges for the RWWP unit processes (Table 1).

Multivariate data analysis

The final CCA results are displayed in Figure 2, indicating the relationship between cyanobacteria and associated organic compounds. Samples (n = 47) were analysed for efficacy of the RWWP (based on the removal of cyanobacteria and associated organic compounds). The cumulative percentage variance of species-environmental composition is explained by correspondence axes 1 and 2, explaining 77.8% and 100% respectively. Microcystin, geosmin and Chl–665 revealed the biggest influence on the data during the spring-summer months. Samples collected during autumn-winter months were closely associated with TOC, while cyanobacteria genera were relatively consistently associated with both seasonal periods (Figure 2).

Figure 2

CCA multivariate statistical results for cyanobacteria genera and associated organic compounds.

Figure 2

CCA multivariate statistical results for cyanobacteria genera and associated organic compounds.

A positive correlation was also observed between Anabaena sp., Chl-665 and geosmin. The positive correlation between Chl-665 and Anabaena sp. indicates that most of the photosynthetic pigments originate from Anabaena sp. TOC was negatively correlated with Anabaena sp., Chl-665 and geosmin. The taste and odour present in the water from the backwash cycle can thus be ascribed to the presence of geosmin produced by Microcystis sp. and Anabaena sp., which is also confirmed by Ewerts et al. (2013), who found a positive correlation between Anabaena sp. and geosmin. Shang et al. (2018) also investigated the co-occurrence of microcystins and taste and odour compounds in drinking water source and their removal in a full-scale drinking water treatment plant. This study confirmed the production of geosmin with 2–methylisoborneol (2-MIB) and other organic compounds such as ß-cyclocitral and ß–ionone.

Cyanobacteria and associated organic compound assessment: after sedimentation

The number of cyanobacteria cells after sedimentation in the RWWP was reduced to an average concentration of 68 cells/mL (Table 3) during the autumn-winter months and an average of 259 cells/mL (Table 3) during spring-summer months. All phytoplankton concentrations were reduced to acceptable levels during the study period; however, the concentrations of organic compounds (TOC and DOC) remained relatively stable throughout the study period. The conventional processes require optimisation strategies to remove these compounds, especially since the raw water from filter backwash is enriched with organic compounds. At this stage of the treatment process, the total phytoplankton composition remained dominated (mean of 50% and above) by cyanobacteria during the study period and showed effective removal for taste and odour compounds, in particular for geosmin concentrations during the spring-winter months. Ewerts et al. (2013) found that the optimisation of coagulation, flocculation and sedimentation can remove a large portion of phytoplankton (which include cyanobacteria) and subsequently avoid the lysis of cells that may increase the content of organic compounds in water.

Cyanobacteria and associated organic compound assessment: after filtration

The number of cyanobacteria cells present after filtration were recorded at <60 cells/mL for the study period. The South African water quality guideline for domestic use (DWAF 1996) prescribes a threshold of 500 cells/mL present at this stage of treatment or in drinking water to limit the risk from exposure to toxins. However, no potential incidents could be expected, since cell concentrations and associated organic compounds were effectively removed to relatively low levels (Table 3). This effective removal of cell quantities subsequently promotes the removal of associated organic compounds such as cyanotoxins. Microcystin concentrations were below the guideline levels of 1 μg/L after filtration, which is also acceptable for drinking water. However, such low concentrations of microcystin can also be ascribed to frequency of occurrence of cyanobacteria cells (generally known for producing microcystin); therefore, continuous sampling and water quality monitoring in order to optimize unit processes should be implemented (Du Preez & Avenant-Oldewage 2007).

ANOVA analyses, percentage removal and average concentrations for sampling points

The presence of cyanobacteria in raw water abstracted for water treatment might affect the quality of water produced and has been found to be potentially toxic to animals and humans (Chinyama et al. 2016). An ANOVA of the percentage removal for the study period was performed to indicate statistically significant differences between the sampling points for cyanobacteria (Microcystis sp., Anabaena sp. and Oscillatoria sp.) and organic compounds. Results (Table 3) gives the average concentrations at each stage of treatment to support statistical outputs (Table 2).

Table 2

Statistical results showing the removal efficiencies for cyanobacteria genera and associated organic compounds from different stages of the RWWP, which include raw water (RW), after sedimentation (AS) and after filtration (AF)

Cyanobacteria genera and associated organic compoundsUnitsStatistically Significant Differences (p-values)
Percentage removal (%)
Cyanobacteria genera 
  autumn-winter spring-summer autumn-winter spring-summer 
Microcystis sp. cells/mL 0.02 ✓ 0.03 ✓ 99.60 99.90 
Anabaena sp. cells/mL 0.11 ✗ 0.01 ✓ 99.60 99.20 
Oscillatoria sp. cells/mL 0.30 ✗ 0.36 ✗ 99.18 99.66 
Organic compounds 
Geosmin ng/L 0.03 ✓ 0.07 ✗ 63.00 85.70 
Chl − 665 μg/L 0.00 ✓ 0.00 ✓ 99.26 99.25 
Microcystin μg/L 0.20 ✗ 0.01 ✓ 58.57 73.45 
TOC mg/L as C 0.44 ✗ 0.95 ✗ 4.62 2.20 
DOC mg/L as C 0.81 ✗ 0.14 ✗ 11.33 29.28 
Cyanobacteria genera and associated organic compoundsUnitsStatistically Significant Differences (p-values)
Percentage removal (%)
Cyanobacteria genera 
  autumn-winter spring-summer autumn-winter spring-summer 
Microcystis sp. cells/mL 0.02 ✓ 0.03 ✓ 99.60 99.90 
Anabaena sp. cells/mL 0.11 ✗ 0.01 ✓ 99.60 99.20 
Oscillatoria sp. cells/mL 0.30 ✗ 0.36 ✗ 99.18 99.66 
Organic compounds 
Geosmin ng/L 0.03 ✓ 0.07 ✗ 63.00 85.70 
Chl − 665 μg/L 0.00 ✓ 0.00 ✓ 99.26 99.25 
Microcystin μg/L 0.20 ✗ 0.01 ✓ 58.57 73.45 
TOC mg/L as C 0.44 ✗ 0.95 ✗ 4.62 2.20 
DOC mg/L as C 0.81 ✗ 0.14 ✗ 11.33 29.28 

The tick symbol ‘✓ ‘ indicates significant difference where p<0.05, while a cross symbol ‘✗‘ indicates no significant difference where p>0.05 and cyanobacteria and organic compounds removal efficiencies in percentage (%).

Table 2 displays the ANOVA results comparing the overall water treatment processes (coagulation, flocculation, sedimentation and filtration). This was performed for cyanobacteria as well as for the organic compounds geosmin, Chl-665, TOC and DOC. The statistical results obtained for cyanobacteria and microcystin displayed statistically significance differences for the study period and there was not enough evidence to suggest any statistical differences for geosmin, Chl-665, TOC and DOC (Tables 2 and 3). To evaluate the individual processes and their capabilities to remove cyanobacteria and associated organic compounds, statistical tests were performed between data collected from each unit process to determine if a statistically significant difference existed between the water quality before and after each unit process. The statistical results show that cyanobacteria and microcystin were removed effectively during coagulation, flocculation and sedimentation as well as during sand filtration. Geosmin was effectively removed by sand filtration and no statistically significant difference was found during the sedimentation process.

The percentage removal by the RWWP for the removal of cyanobacteria (Microcystis sp., Anabaena sp. and Oscillatoria sp.) were up to 99% efficiency. The results show that the presence of cyanobacteria did not have a major impact on the water treatment processes although initial concentrations were recorded in relatively abundant quantities (78,192 cells/mL maximum). Removal of these cells is a priority before recovered water is recycled back to the final treated water of the main treatment works, especially since the concentrations in the recovered water still exceed the acceptable concentration of 2,000 cells/mL recommended by the World Health Organization (WHO) guidelines (WHO 1998). The highest concentrations (average concentrations of 6,103 cells/mL and 18,310 cells/mL for autumn-winter and spring-summer months respectively) was due to Microcystis sp., a genus known to produce microcystin (Tanabe et al. 2018). Anabaena sp. was detected only once in the raw water during the autumn-winter months. The filamentous Oscillatoria genus occurred on only a few occasions and was not present at all sampling localities; the low frequency of occurrence of this filamentous genus in raw and after filtration of this treatment works is supported by Ewerts et al. (2013).

The removal efficiencies recorded for the autumn-winter and spring-winter months were 63% and 85.7% respectively and overall removal efficiency was 83.3%. The odour causing compound geosmin was reduced to acceptable levels after filtration prior to recycling water to the main treatment works. Geosmin was sporadically detected in filter backwash water treated by the RWWP, mostly during spring-summer months with occasional concentrations exceeding the production guideline of 30 ng/L of the main treatment works. This was most probably due to the succession of cyanobacteria species producing the different organic compounds. Geosmin can be expected in the water when Anabaena sp. cells are present either in moderate or abundant quantities. Geosmin is usually only removed by advanced treatments like adsorption to powdered activated carbon, granular activated carbon, or oxidation by ozone when released into the water (Westerhoff et al. 2005). Therefore, although taste and odour compounds or other organic compounds may enter the final stage or water treatment, chlorination or disinfection will reduce compounds to acceptable levels. Chlorination is one of the methods, widely applied for various water treatment purposes, especially to control algal growth and oxidize organic material (Van der Walt et al. 2009).

The Chl-665 in the RWWP raw water samples during spring-summer months exceeded the maximum recommended limit of (0–15 μg/L), but the plant managed to remove Chl-665 effectively. The Chl-665 contents in recovered water were effectively removed during the study period. The percentage removal of Chl-665 was recorded up to 99%. High levels of chlorophyll indicate high levels of phytoplankton cells in water (Ewerts et al. 2015), which may occur as result of algal growth and blooms when nutrients such as nitrogen and phosphorus are entering surface water destined for drinking water production. According to Ewerts et al. (2015), Chl-665 analysis can be useful in selecting coagulant treatment that will improve the removal efficiencies of phytoplankton and reduce treatment costs.

The percentage removal of microcystin ranged from 58.6% during autumn-winter and 73.5% during spring-summer months, and overall 67% removal during the study period. Microcystin concentrations on average during spring-summer months were higher than the recommended level of 1 μg/L for raw water. TOC removal percentage was low at 4.6% during autumn-winter months and 2.2% during spring-summer months, and 3.5% removal efficiency over the study period. The most effective removal occurred during spring-summer months for DOC, also showing a removal percentage of up to 29.3% (overall study period 21.0% removal). When organic compounds (TOC and DOC) react with chlorine during the disinfection process, the formation of harmful by-products may pose a risk to drinking water consumers. It is thus important for organic compounds to meet the required standards for drinking water. South African National Standards 241 (SANS 2011) prescribe a chronic health limit (≤10 mg/l) for TOC, which include phytoplankton and chlorophyll. In the distribution network, increased organic compounds may lead to the potential growth of microorganisms (Ozdermir 2014).

Table 3 below indicate the average concentration of phytoplankton and associated organic compounds measured at different sampling points during seasonal study periods (autumn-winter and spring-summer months). The average values were used to calculate percentage removal according to the formula described in the methodology. Most of the parameters decreased with the flow of water treatment, indicating efficacy of the RWWP, except for TOC during autumn-winter months and measurements remained constant during spring-summer months. Table 2 above also confirms the removal efficacy of parameters.

Table 3

The phytoplankton and associated organic compound average concentrations measured in the raw water (RW) and at different stages (AS-AF) during seasonal periods of the study

Autumn-winter months
Spring-summer months
RWASAFRWASAF
ParametersMeasuring unitsAverage concentrations
Average concentrations
Bacillariophyceae cells/mL 769 37 909 23 
Chlorophyceae cells/mL 200 22 315 
Cryptophyceae cells/mL 291 16 227 
Dinophyceae cells/mL 52 40 
Euglenophyceae cells/mL 45 40 
Cyanophyceae cells/mL 6,593 68 26 22,653 259 58 
Microcystis sp. cells/mL 6,103 15 18,310 83 33 
Anabaena sp. cells/mL 199 46 23 4,358 171 24 
Oscillatoria sp. cells/mL 291 40 
TOC mg/l as C 
DOC mg/l as C 
Geosmin ng/l 59 32 
Chlorophyll − 665 μg/l 18 34 
Microcystin μg/l 
Autumn-winter months
Spring-summer months
RWASAFRWASAF
ParametersMeasuring unitsAverage concentrations
Average concentrations
Bacillariophyceae cells/mL 769 37 909 23 
Chlorophyceae cells/mL 200 22 315 
Cryptophyceae cells/mL 291 16 227 
Dinophyceae cells/mL 52 40 
Euglenophyceae cells/mL 45 40 
Cyanophyceae cells/mL 6,593 68 26 22,653 259 58 
Microcystis sp. cells/mL 6,103 15 18,310 83 33 
Anabaena sp. cells/mL 199 46 23 4,358 171 24 
Oscillatoria sp. cells/mL 291 40 
TOC mg/l as C 
DOC mg/l as C 
Geosmin ng/l 59 32 
Chlorophyll − 665 μg/l 18 34 
Microcystin μg/l 

CONCLUSIONS

The study investigated the efficacy of RWWP in removing phytoplankton and associated organic compounds. The results show that the recycled raw water supplied to RWWP contained a phytoplankton composition consisting of six major groups (diatoms, green algae, cyanobacteria, dinoflagellates, golden-brown algae and euglenoids). Among others, the cyanobacteria group (consisting of Anabaena sp., Microcystis sp., and Oscillatoria sp.) was the most dominant group (mean of 80% and above). Although the RWWP uses conventional unit processes, it was able to reduce the dominant cyanobacteria group effectively (up to 99% removal) during different seasonal occasions. Chlorophyll (Chl-665) removal percentages were relatively similar to percentages recorded for cyanobacteria, which was expected since this organic compound can be measured as a direct indication of phytoplankton content in water. Other organic compounds, which include geosmin, microcystin, TOC and DOC, overall removal percentages removal were lower (varying from 2.20% to 85.70%). The water treatment processes implemented in the RWWP require optimization strategies to improve the removal of organic compounds, especially for TOC and DOC. Cyanobacteria present in recycled water destined for treatment by the RWWP don't pose a risk to drinking water quality; however, chemical and operational optimization practices could improve the removal efficiencies of all organic compounds prior to mixing treated recycled water with treated water in the main treatment plant.

RECOMMENDATIONS

  • Implement phytoplankton water quality monitoring for RWWP. Phytoplankton species, especially those belonging to the cyanobacteria group, are known to be complex microorganisms due to cell morphology, size and organelles such as gas vacuoles. The characteristics of cell, colonies or filaments give organisms the ability to maintain their position in the water column; thus, the effective removal of various phytoplanktons should for an integral part of water quality monitoring practices where the treatment of recycled wash water is applied.

  • Implement a routine sampling plan to determine water quality at different stages of the RWWP. Routine sampling practices are often scheduled for the main treatment plant; however, different stages of water treatment in recycled water plants should be incorporated in routine sampling plans where the capacity exist. The treatment and monitoring costs associated with recycled water treatment plants may be significantly higher due to accumulation of impurities in sand filters. Occasionally, the operations of recycling water treatment may also be discontinued, depending on seasonal changes or main treatment plant operations. Therefore, effective management (budgeting and planning) may add significant value to treatment facilities where water shortages or water balance issues are posing risks to meet the demand for drinking water by consumers.

  • Consider stringent water quality monitoring plans to be implemented during all seasonal periods in the RWWP. The presence of high concentrations of potentially harmful organic compounds (e.g. cyanotoxins) as well as other organic compounds (e.g. organic carbon) known as precursors for the formation of disinfection by-products were detected in the RWWP. Each step of treatment should be carefully monitored and guided by stringent measures to avoid contamination of water in the main treatment works.

ACKNOWLEDGEMENTS

Rand Water Board is acknowledged for their facilities, infrastructure and financial support to the project. Family, friends and colleagues are thanked for moral support.

REFERENCES

REFERENCES
APHA (American Public Health Association)
2013
Standard Methods for the Examination of Water and Wastewater
, 22nd edn.
APHA
,
Washington, DC
, p.
733
.
Chinyama
A.
,
Ochleng
G. M.
,
Snyman
J.
&
Nhapi
I.
2016
Occurrence of cyanobacteria genera in the Vaal Dam: implications for potable water production
.
Water SA
42
(
3
),
415
420
.
Chorus
I.
&
Bartram
J.
1999
Toxic Cyanobacteria in Water: A Guide to Their Public Health Consequences, Monitoring and Management
.
World Health Organisation
.
Available from: https://apps.who.int/iris/handle/10665/42827 (accessed 23 April 2019).
Du Preez
H.
&
Avenant-Oldewage
A.
2007
Cyanobacterial incident management frameworks (CIMFS) for application by drinking water suppliers
.
Water SA
33
(
5
),
643
652
.
DWAF (Department of Water Affairs and Forestry)
1996
South African Water Quality Guidelines, Volume 1: Domestic Water use
, 2nd edn.
Department of Water Affairs and Forestry
,
Pretoria
,
South Africa
.
DWAF (Department of Water Affairs and Forestry)
2002
Using Water Wisely: A National Water Resource Strategy for South Africa Information Document
.
Pretoria
,
South Africa
.
Envirologix Quantiplate Kit for Microcystins
2013
Catalog No.: EP 022, 500 Riverside Industrial Parkway, Portland, ME 04103-1486, USA. 7
.
Ewerts
H.
,
Swanepoel
A.
,
Du Preez
H. H.
&
Van der Walt
N.
2015
Total photosynthetic pigments in addition to turbidity during the selection of coagulant treatments: a drinking water treatment perspective
.
Journal of Water Supply: Research and Technology
64
(
1
),
1
9
.
Govender
J.
2016
An Assessment of the Water Quality of the Baynespruit River and its Linkages to the Health of the Sobantu Community
.
Master of Science dissertation
,
University of Kwazulu-Natal
,
Pietermaritzburg
,
South Africa
Grey
D.
&
Sadoff
C. W.
2007
Sink or Swim?: water security for growth and development
.
Water Policy
9
(
6
),
545
571
.
Harding
W. R.
&
Paxton
B. R.
2001
Cyanobacteria in South Africa: A Review
.
Water Research Commission Report TT 153/01
,
Pretoria
,
South Africa
.
Hitzfeld
B. C.
,
Hogers
S. J.
&
Dietrich
D. R.
2000
Cyanobacterial toxins: removal during drinking water treatment, and human risk assessment
.
Environmental Health Perspective
108
(
1
),
113
122
.
Li
W.
,
Liang
X.
,
Duan
J.
,
Beecham
S.
&
Mulcaby
D.
2017
Influence of spend filter backwash water recycling on pesticide removal in a conventional drinking water treatment process
.
Environmental Science: Water Research & Technology
4
,
1057
1067
.
Matthews
M. W.
&
Bernard
S.
2015
Eutrophication and cyanobacteria in South Africa standing water bodies: a view from space
.
South African Journal of Science
111
(
5/6
),
1
8
.
Oberholster
P. J.
2010
A CSIR perspective on water in SA: the current status of water quality in SA
.
Available from: www.csir.ac.za (accessed 16 May 2015)
.
Oberholster
P. J.
,
Botha
A. M.
&
Grobbelaar
J. U.
2004
Microcystis aeruginosa: source of toxic microcystins in drinking water
.
African Journal of Biotechnology
3
(
3
),
159
168
.
Shang
L.
,
Feng
M.
,
Xu
X.
,
Liu
F.
,
Ke
F.
&
Li
W.
2018
Co-occurrence of Microcystins and taste-and-odor compounds in drinking water source and their removal in a full-scale drinking water treatment plant
.
Toxins
10
(
26
),
1
17
.
Sibande
R. X.
2013
Cost of Eutrophication at the Vaal River System: An Integrated Economic Model
.
Master of Commerce Dissertation
,
University of Pretoria
,
Pretoria
,
South Africa
.
South African National Standards 241-1 (SANS 214-1)
2011
Drinking Water. Part 1: Microbiological, Physical, Aesthetic and Chemical Determinants
, 1st edn.
SABS Standards Division, Pretoria
,
South Africa
.
Swanepoel
A.
2015
Early Warning System for the Prediction of Algal-Related Impacts on Drinking Water Purification
.
PhD Thesis
,
North west University
,
Potchefstroom
,
South Africa
, p.
152
.
Swanepoel
A.
,
Du Preez
H. H.
,
Schoeman
C.
,
Janse Van Vuuren
S.
&
Sundram
A.
2008
Condensed Laboratory Methods for Monitoring Phytoplankton, Including Cyanobacteria, in South African Freshwaters
.
Water Research Commission Report TT 323/08
,
Pretoria
,
South Africa
.
Tanabe
Y.
,
Hodok
Y.
,
Sano
T.
,
Tada
K.
&
Watanabe
M. M.
2018
Adaptation of the freshwater bloom-forming cyanobacterium microcystis aeruginosa to brackish water is driven by recent horizontal transfer of sucrose genes
.
Fronties in Microbiology
9
,
1
11
.
Van Der Walt
M.
,
Krüger
M.
&
Van der Walt
C.
2009
The South African Oxidation and Disinfection Manual. WISA Oxidation and Disinfection Division
.
WRC Report No. TT 406/09
.
Water Research Commission
,
Pretoria
.
Van Vuuren
S. J.
,
Taylor
J.
,
Van Ginkel
C.
&
Gerber
A.
2006
Easy Identification of the Most Common Freshwater Algae: A Guide for the Identification of Microscopic Algae in South African Freshwaters
.
Department of Water Affairs and Forestry and North-West University
,
North West, Potchefstroom, South Africa
,
ISBN:0-621-35471-6
.
WEF (World Economic Forum)
2017
The Global Risks Report 2017
, 12th edn.
Insight Report. ISBN: 978-1-944835-07-1 (REF: 050117)
,
Geneva
,
Switzerland
.
Westerhoff
P.
,
Summers
R. S.
,
Chowdhury
Z.
&
Kommineni
S.
2005
Ozone-enhanced Biofiltration for Geosmin and MIB Removal
.
AWWA Research Foundation
,
Denver
,
1
220
.
World Health Organization (WHO)
1998
Guidelines for Drinking Water Quality Addendum Volume 2, Cyanobacterial Toxins: Microcystin L.R
.
World Health Organization
,
Geneva, Switzerland
.