Abstract
Lake Chivero, Harare's major raw water source, is eutrophic and algae-infested. Consequently, the conventional water treatment processes (CWTPs) that are used at Harare's main water treatment works, Morton Jaffray Water Treatment Works (MJWTW), have been ineffective in algae removal. This study was carried out to investigate the feasibility of using dissolved air flotation (DAF) to remove algae at MJWTW. Experiments were carried out using a 60-L DAF pilot system with a hydraulic loading rate of 7 m/h. Raw and treated water was characterised in terms of turbidity, pH and electrical conductivity (EC), and chlorophyll-a. Results showed reductions in turbidity, pH, and chlorophyll-a of 64, 27, and 95%, respectively. However, EC increased by 42% due to the addition of alum and acid. The DAF performed better than the CWTP currently used at MJWTW based on a comparison to a study carried out in the same period simulating conventional treatment of the same raw water and values reported in the literature for the CWTP. The adoption of DAF at MJWTW could potentially result in more efficient water treatment and better algae removal too. Further tests on optimum acid dosing for DAF are recommended while using alum as a coagulant.
HIGHLIGHTS
Lake Chivero is eutrophic and algae-infested.
The current conventional water treatment process (CWTP) being used are ineffective in algae removal.
As such, water treatment costs at Morton Jaffray Water Treatment Works (MJWTW) have increased.
Thus, algae have been observed in Harare's water distribution system.
Dissolved air flotation (DAF) was considered an alternative to the CWTP as it is more effective in algae removal.
Graphical Abstract
INTRODUCTION
Globally, the pollution of water resources has led to a decrease in water quality and rapid algae growth, which is associated with numerous health problems in humans (Czyżewska & Piontek 2019). As such, algae must be removed during the water treatment process. Algae can be removed by microstraining (physical), algaecides (chemical), and the application of coagulation, flocculation, sedimentation, filtration, and disinfection (conventional) respectively (Czyżewska & Piontek 2019; Shen et al. 2019). Although conventional methods are the most widely used techniques in algae removal in developing countries, they are now failing to deal with high levels of algae pollution (Hoko et al. 2021b). Conventional water treatment processes (CWTPs) fail to adequately remove algae since several algae species can regulate their buoyancy when alive and thus can avoid settling as stated by Hoko & Makado (2011). This presents a challenge when the CWTP is used to remove algae. Consequently, high doses of coagulant need to be applied for the effective removal of algae (Ghernaout et al. 2010). Therefore, dissolved air flotation (DAF), which takes advantage of the natural buoyancy of algae, can be applied to treat algae-infested water (Edzwald 2006).
DAF is a process that uses microbubbles to remove suspended matter dispersed in a liquid (Crittenden et al. 2012). DAF is the most suitable method for low-density particles and flocs (Edzwald 2006). The bubbles formed rise to the surface of the water with flocs attached to them due to their low density and are later skimmed off at the top (Edzwald 2011). After the water has been treated by DAF, clarified water (subnatant or floated water) is removed from the bottom while the accumulated sludge (float) is removed from the top (Brandt et al. 2017). The DAF process floats particles using microscopic bubbles that are produced by saturating recycled water with air at high pressure and subsequently releasing it at atmospheric pressure (Henderson et al. 2009). This makes DAF efficient for algae removal as the sludge with algae is raised to the surface. The sludge is removed by mechanical or hydraulic means (Crittenden et al. 2012). The most commonly applied mechanical sludge removal methods are scrappers which consist of rubber blades that push sludge into the collection channel and beach scrapers which scrap sludge from the sludge beach (Zabel 1985). DAF results in a 30% reduction in the amount of coagulant used compared to the CWTP for algae removal (Khiadani et al. 2014).
DAF has a cost–benefit extremely advantageous when compared to most algae removal techniques including centrifugation, filtration, flocculation, sedimentation, and flotation, demanding just 0.015 kW/m³ (Torres et al. 2017). DAF has been widely used to remove algae from water and has often proven more efficient than sedimentation for algae–water separation (Zhang et al. 2014). Teixera & Rosa (2007) concluded that DAF had better results as compared to clarifiers which resulted in reduced algal load on the filters consequently leading to a more efficient filtration process for the treatment of raw water from Thames River in the United Kingdom. Therefore, DAF is cost-effective and efficient for algae removal. However, the advantages of algae removal using DAF have not been fully exploited in southern Africa, where CWTPs are commonly used although the technology has been used in South Africa as stated by Muñoz-Alegría et al. (2021). An example of an algae-infested raw water source for drinking water is Lake Chivero in Zimbabwe (Dandadzi et al. 2020).
Lake Chivero has been reported to be highly eutrophic since the 1960s and its water quality has been getting worse with time (Nhongo et al. 2018). Consequently, Lake Chivero's water is now algae-infested, resulting in water treatment problems at Harare's main water treatment works, Morton Jaffray Water Treatment Works (MJWTW), as stated by Hoko et al. (2021b). The presence of algae in Lake Chivero has caused unpleasant taste and odour problems in drinking water in Harare and has also led to the detection of algae in Harare's water distribution system (Dandadzi et al. 2019). This implies that the CWTPs currently applied at MJWTW are ineffective in algae removal. Consequently, this has led to multiple challenges, including a low willingness to pay for water by the City of Harare's customers. It is for this reason that this study investigated the feasibility of the application of the DAF at MJWTW for algae removal, through laboratory experiments. The study was carried out in the period from April to May 2014. Physical (turbidity), chemical [pH and electrical conductivity (EC)], and biological (chlorophyll-a) parameters were used to assess the performance of the DAF pilot system.
MATERIALS AND METHODS
Study area
Location of study site
Background on MJWTW
Lake Chivero has become very eutrophic, thus algae have become one of the main problems affecting treatment of potable water at MJWTW (Hoko et al. 2021b). As such, the filter backwashing frequency has been between 4 and 8 h against the recommended 24–36 h and this has resulted in inefficient turbidity removal and low plant output (Hoko & Makado 2011). Consequently, there are recurring problems related to the quality and quantity of drinking water in Harare that have been linked to repeated outbreaks of cholera (WHO 2018).
Study design
Selection of study site
Harare, the capital city of Zimbabwe, faces drinking water quality problems because Lake Chivero is eutrophic (Hoko et al. 2021b). Consequently, MJWTW has been failing to produce acceptable-quality water (Dandadzi et al. 2019). There has been evidence of poor water quality in Harare's distribution system, including algae regrowth (Dandadzi et al. 2020). Thus, an effective algae removal process needs to be adopted at MJWTW.
Selection of study parameters
Physical (turbidity), chemical [pH and EC], and biological (chlorophyll-a) parameters were considered for this study. Turbidity is one of the key parameters for monitoring water and affects aesthetics (Sawyer et al. 2003). The pH of water affects the disinfection process, especially when chlorine is used, as is the case at MJWTW (Hoko & Makado 2011). EC is used to monitor the performance of water treatment (Sawyer et al. 2003). The presence of chlorophyll-a in water implies the presence of algae (EPA 2022).
Experimental design
A laboratory DAF pilot system was designed and set up in a laboratory at the University of Zimbabwe to assess the feasibility of the application of DAF for water treatment at MJWTW. The fabrication of the DAF pilot plant was followed by experiments using raw water obtained from the inlet at MJWTW. To compare DAF with the CWTP, the results of the DAF system were compared with those of a study by Hoko et al. (2021a) carried out during the same period using raw water from MJWTW. The study by Hoko et al. (2021a) investigated the use of alternative coagulants applied to the CWTP.
Design and layout of the DAF pilot plant
Figure 3 shows the layout of the DAF pilot system.
Layout of the DAF basin
DAF pilot system design and operational parameters
Table 1 shows the DAF pilot system design parameters which were adopted for this study.
Parameter . | Range in the literature . | Selected value . | Remark . |
---|---|---|---|
Hydraulic loading rate (HLR) (m/h) | 5–15a | 5–10 | Based on the available pump |
Recycle rate (%) | 5–12b | 5–10 | – |
Pressure (kPa) | 300–600 | 300 | – |
Process inflow rate (L/s) | – | 0.2 | Based on the available pump |
Surface area (m2) | – | 0.2 | Based on HLR |
Retention time (min) | 5–15c | 5 | Based on the volume |
Parameter . | Range in the literature . | Selected value . | Remark . |
---|---|---|---|
Hydraulic loading rate (HLR) (m/h) | 5–15a | 5–10 | Based on the available pump |
Recycle rate (%) | 5–12b | 5–10 | – |
Pressure (kPa) | 300–600 | 300 | – |
Process inflow rate (L/s) | – | 0.2 | Based on the available pump |
Surface area (m2) | – | 0.2 | Based on HLR |
Retention time (min) | 5–15c | 5 | Based on the volume |
Design of the flocculation basin
Determination of the coagulant dosage
Jar tests were carried out to determine the optimum alum dosage as recommended by Langlais & Degremont (2010). Alum was used because it is the coagulant currently used at MJWTW (Hoko & Makado 2011). Before the jar tests were carried out, the pH was lowered using sulphuric acid as is the practice at MJWTW to ensure that the pH falls within the optimum pH range for alum (Hoko et al. 2021b).
Operation of the DAF pilot system
The operation of the DAF system involved turning on the inflow pump and setting the inflow to the flocculation basin to the desired flow. The coagulant dosing was performed in the flocculation basin and a mechanical mixer was used for the rapid mixing process. The pH of the raw water was lowered using sulphuric acid to the recommended range for alum of around 6–7 (EPA 2017). An alum dosage of 12.46 mL/s was derived based on the dose (70 mg/L) determined from the jar tests, the process flow rate, and the percentage of aluminium oxide (Al2O3) in the alum. Rapid and slow mixing occurred for 5 min (alum reaction) and 2 min (flocculation) respectively in the flocculation basin before the water being fed to the DAF pilot system as recommended by Edzwald (1995). A saturator, which was elevated above the DAF basin to prevent backflow of water, was used to promote the formation of small bubbles. Air was blown into the system using a compressor through 1-mm nozzles as soon as the water was in the contact zone of the DAF basin. The sizes of the air bubbles were altered by adjusting the air inlet valve on the saturator. An air control valve and a pressure gauge were incorporated to allow regulation of air and measurement of the pressure. Recycling was initiated at the instance when the water level in the DAF basin was above the recycle pump suction level. The pressure in the saturation tank was controlled using a pressure regulator with a pressure of around 300–400 kPa as recommended by Liu et al. (2007). The recycling rate was also adjusted to balance the incoming air and the recycled water in the saturator using a calibrated valve on the saturator. The scum was removed by raising the water level in the flotation tank by throttling the outlet valve to allow the sludge to flow into the scum collection chamber. A pump was connected to recycle the water into the saturation vessel. The recycling rate was maintained between the recommended range of 6 and 12% (Edzwald 2011). The treated water samples were collected at the outlet for characterisation. The experiments were carried out to determine optimum pressure, coagulant dosage, and recycle ratio. Seven experiments were carried out for 30 min each (after adjustment of all operational parameters to ensure a steady state performance of the system) over 5 weeks, i.e. five campaigns.
The CWTP design and operation
This section describes the CWTP design and operation outlined in the reference study by Hoko et al. (2021a). The study was a parallel study carried out at the same time using the same raw water source in the same laboratory. The text below is extracted from Hoko et al. (2021a). The CWTP pilot plant comprised of a flocculation basin and sedimentation tank which were setup and fabricated in the University of Zimbabwe Construction and Civil Engineering Department Hydraulics laboratory. The setup was to test the performance of selected alternative coagulants having alum as the control, under flow through conditions to mimic the process at MJWTW.
Flocculation basin design
Parameter . | Unit . | Design value . | Literature ranges . | Literature reference . |
---|---|---|---|---|
t | min | 15 | 10–20 | DWD (1995); Degremont (2007) |
Q | m3/h | 0.1 | – | |
D | m | 0.2 | – | |
As | m2 | 0.0125 | – |
Parameter . | Unit . | Design value . | Literature ranges . | Literature reference . |
---|---|---|---|---|
t | min | 15 | 10–20 | DWD (1995); Degremont (2007) |
Q | m3/h | 0.1 | – | |
D | m | 0.2 | – | |
As | m2 | 0.0125 | – |
Parameter . | Unit . | Value . |
---|---|---|
Length | m | 0.6 |
Width | m | 0.15 |
Depth | m | 0.2 |
Area | m2 | 0.0125 |
Parameter . | Unit . | Value . |
---|---|---|
Length | m | 0.6 |
Width | m | 0.15 |
Depth | m | 0.2 |
Area | m2 | 0.0125 |
where V is the volume of the basin (m3), L is the length (m), W is the width (m), D is the depth (m).
Rectangular sedimentation tank design
Parameter . | Unit . | Design . | Literature ranges . | Literature references . |
---|---|---|---|---|
vs | m3/m2·h | 0.5 | 0.5–1 | Degremont (2007) |
t | h | 0.5 | 2–4 | Bahadori et al. (2013) |
D | m | 0.25 | – | – |
Parameter . | Unit . | Design . | Literature ranges . | Literature references . |
---|---|---|---|---|
vs | m3/m2·h | 0.5 | 0.5–1 | Degremont (2007) |
t | h | 0.5 | 2–4 | Bahadori et al. (2013) |
D | m | 0.25 | – | – |
SLR, surface loading rate; t, hydraulic detention time; As, surface area of basin; D, tank depth.
Parameter . | Unit . | Value . |
---|---|---|
Length | m | 0.90 |
Width | m | 0.20 |
Depth | m | 0.25 |
As | m2 | 0.2 |
Parameter . | Unit . | Value . |
---|---|---|
Length | m | 0.90 |
Width | m | 0.20 |
Depth | m | 0.25 |
As | m2 | 0.2 |
As, surface area of basin.
Operation of the CWTP plant
Raw and treated water sampling
Seven composite raw water samples were obtained from Lake Chivero on different days and treated in the DAF system each time samples were collected. They were transported to the laboratory in 20-L plastic containers. From these containers, the water was taken and fed into the DAF pilot system. Grab samples for treated water were collected on the lower outlet of the DAF system.
Methods of water quality analysis
Raw and treated water was characterised in terms of turbidity, pH, EC and chlorophyll-a concentration. The characterisation was carried out at the Department of Construction and Civil Engineering and the Department of Biological Sciences laboratories of the University of Zimbabwe. EC, pH and turbidity were analysed according to the procedures outlined in Standard Methods for the Examination of Water and Wastewater (APHA 2012). Chlorophyll-a concentration was determined using the method described by Brönmark & Hansson (2005). Table 6 shows the equipment and methods used for water quality testing.
Parameter . | Analytical instrument . | APHA method number . | Equipment brand . |
---|---|---|---|
Turbidity | Turbidity meter | 2130B | Hanna (HI 9803) |
pH | pH ion meter | 4500-H+ | Hanna |
Electrical conductivity | Conductivity meter | – | Lasany Microprocessor 1–50 |
Chlorophyll-a | Acetone extraction | – | – |
Parameter . | Analytical instrument . | APHA method number . | Equipment brand . |
---|---|---|---|
Turbidity | Turbidity meter | 2130B | Hanna (HI 9803) |
pH | pH ion meter | 4500-H+ | Hanna |
Electrical conductivity | Conductivity meter | – | Lasany Microprocessor 1–50 |
Chlorophyll-a | Acetone extraction | – | – |
Methods of data analysis
Statistical analysis that included the coefficient of variation (CV) and t-test were used to determine the variation of the water quality parameters and also the efficiency of water treatment by DAF.
RESULTS AND DISCUSSION
Raw water characteristics
Table 7 shows the results for raw water characterisation of the Lake Chivero water which was collected at MJWTW inlet works.
Campaign number . | 1 . | 2 . | 3 . | 4 . | 5 . | 6 . | 7 . | Mean . | Std deviation . | CV (%) . |
---|---|---|---|---|---|---|---|---|---|---|
Turbidity (NTU) | 4.04 | 4.04 | 3.72 | 3.72 | 3.14 | 3.26 | 3.26 | 3.60 | 0.38 | 10.5 |
pH | 8.0 | 8.4 | 7.3 | 7.7 | 8.0 | 7.2 | 7.5 | 7.73 | 0.43 | 5.6 |
EC (μS/cm) | 483 | 483 | 328 | 328 | 339 | 339 | 339 | 377 | 72.6 | 19.3 |
Chlorophyll-a (μg/L) | 2.12 | 2.12 | 1.92 | 1.92 | 0.36 | 0.36 | 0.36 | 1.31 | 0.89 | 68.1 |
Campaign number . | 1 . | 2 . | 3 . | 4 . | 5 . | 6 . | 7 . | Mean . | Std deviation . | CV (%) . |
---|---|---|---|---|---|---|---|---|---|---|
Turbidity (NTU) | 4.04 | 4.04 | 3.72 | 3.72 | 3.14 | 3.26 | 3.26 | 3.60 | 0.38 | 10.5 |
pH | 8.0 | 8.4 | 7.3 | 7.7 | 8.0 | 7.2 | 7.5 | 7.73 | 0.43 | 5.6 |
EC (μS/cm) | 483 | 483 | 328 | 328 | 339 | 339 | 339 | 377 | 72.6 | 19.3 |
Chlorophyll-a (μg/L) | 2.12 | 2.12 | 1.92 | 1.92 | 0.36 | 0.36 | 0.36 | 1.31 | 0.89 | 68.1 |
The turbidity of the raw water ranged from 3.14 to 4.04 NTU with a mean of 3.60 NTU and a low CV of 10.8%. The range of raw water turbidity was comparable to the range of 3.0–3.6 NTU that Hoko et al. (2021a) obtained for the same raw water source during a study carried out around the same time as the current one. This was expected as both studies were carried out around the same time. The turbidity for Lake Chivero raw water was in the lower band for typical surface water sources which normally range from 1 to 200 NTU as highlighted by Sharma & Bhattacharya (2017). Thus, the turbidity of raw water found in this study was in the suggested range for lakes.
The raw water pH ranged from 7.2 to 8.4 with a mean of 7.7. The low CV (5.6%) implied that the pH did not vary much during the study period. The pH range of raw water obtained for this study was comparable to 7.3–8.2 found by Dandadzi et al. (2019) for the same water source. The pH was in the 5.0–9.0 range suggested by EPA (2001) for surface water sources. According to EPA (2017), the optimum pH range for coagulation and flocculation when alum is used as the coagulant is 6–7. Thus, the pH was higher than that required for the coagulant (alum) used for the experiments. Therefore, a pH adjustment was necessary for the effective use of alum.
The raw water EC ranged from 328 to 483 μS/cm with a mean of 377 μS/cm and moderate CV of 19.3% suggesting a moderate variation in the EC of raw water during the study period. The range of EC obtained in this study for raw water was comparable to the range 440–528 μS/cm obtained by Hoko et al. (2021b) for the same water source. The EC obtained in this study was below the 1,000 μS/cm limit highlighted by EPA (2001) for surface water sources. Therefore, the raw water EC was moderate during the study period.
The raw water chlorophyll-a concentration ranged from 0.36 to 2.12 μg/L with a mean of 1.31 μg/L and a high variability (CV = 67.9%). The mean chlorophyll-a concentration in raw water was comparable to 2.28 μg/L obtained by a study by Hoko et al. (2021a) for the same raw water source. The chlorophyll-a concentration in the raw water was lower than the 33.89 μg/L obtained by Mogakabe & Van Ginkel (2008) for Bospoort Dam in South Africa. In addition, the chlorophyll-a concentration was below the 5 μg/L highlighted by ANZECC & ARMCANZ (2000) for freshwater reservoirs and lakes. Dandadzi et al. (2019) stated that the detection of chlorophyll-a infers the presence of algae. According to EPA (2022), one of the symptoms of degraded water quality condition is the increase of algae biomass as measured by the concentration of chlorophyll-a. Therefore, the long-term concentration of chlorophyll-a in the raw water signifies that the raw water is eutrophic. Thus, it can be concluded that there were algae in the raw water from Lake Chivero.
Raw water from Lake Chivero was characterised by moderate turbidity, slightly alkaline pH, high EC and high chlorophyll-a concentration. This indicates that the raw water from Lake Chivero was polluted which could render conventional water treatment ineffective to treat the water. Due to the slightly alkaline pH, it was necessary to adjust the pH of the raw water by adding sulphuric acid for effective coagulation.
Determination of alum dose
Performance of the dissolved air floatation system
Results from the experiments carried out on the treatment of raw water from Lake Chivero using the DAF pilot system are presented in this section. The results are based on the analysis of composite samples of raw water fed into the DAF pilot system and effluent from the system for each of the seven campaigns. The performance of the DAF pilot system for the parameters selected (turbidity, pH, EC, and chlorophyll-a concentration) is presented in the sections that follow. The performance of the DAF pilot system which simulated the processes of coagulation, flocculation, and DAF (termed DAF in this paper) were compared to a similar study carried out during the same period using the Lake Chivero raw water which simulated CWTPs of coagulation, flocculation, and sedimentation. Thus, essentially, the DAF basin should be compared to the sedimentation basin.
Turbidity
The treatment of raw water by the DAF system reduced turbidity by 64%, which was lower than the 82.8% obtained by Teixeira & Rosa (2007) for DAF. This was higher than the 53% found by Hoko et al. (2021a) when the CWTP (across the sedimentation unit) was simulated for the same water source around the same period. However, Officer et al. (2001) obtained a much higher turbidity removal efficiency of 97% when DAF was used at Farmoor Water Treatment Works in the United Kingdom where the raw water from the Thames River was used. To increase the turbidity removal efficiency for the raw water from Lake Chivero, the use of alternative coagulants as suggested by Hoko et al. (2021a) together with DAF could be adopted. It can be concluded that DAF has a higher turbidity removal efficiency for the treatment of raw water from Lake Chivero compared to the CWTP. As such, the City of Harare should consider incorporating DAF in the water treatment process at MJWTW to replace the current sedimentation process being used. This could potentially result in longer filter runs and reduction in the coagulant dose applied during water treatment.
pH
The treatment of water with DAF further reduced the pH by 27% after the pH had initially been reduced due to the addition of sulphuric acid, which is higher than the 10% reduction in pH obtained by Hoko et al. (2021a) where the CWTP was simulated. The pH of water affects the effectiveness of chlorine disinfection, and according to WHO (2011), chlorine disinfection is more effective at low pH values of less than 8. Thus, further tests on determining the optimum acid dosing while using DAF should be carried out using raw water from Lake Chivero.
Electrical conductivity
The treatment of water by DAF increased the EC by 42%. This was higher than the 5% increase obtained by Hoko et al. (2021a) where the CWTP was simulated. The EC increment was likely to have been caused by the addition of the coagulant (alum) for turbidity removal as mentioned by Hoko et al. (2021a). The increase in EC was also likely to have been caused by the addition of sulphuric acid as metals tend to be more soluble when the pH decreases as highlighted by Zhang et al. (2018). Therefore, the adoption of DAF at MJWTW could result in an increase in the EC of the treated water.
Chlorophyll-a concentration
Treatment with DAF resulted in a chlorophyll-a removal efficiency of 95%, similar to the chlorophyll-a removal efficiency of 91.8% obtained by Teixeira & Rosa (2007) when DAF was used to treat water. The chlorophyll-a removal efficiency by the CWTP in a study carried out by Hoko et al. (2021a) using the same source of water was 52%. Dandadzi et al. (2020) established algae regrowth in the water supplied by the City of Harare to its residents. Therefore, it was concluded that DAF had a higher chlorophyll-a removal efficiency compared to the CWTP. The adoption of DAF in the water treatment process as MJWTW could significantly reduce or eliminate algae in treated water.
CONCLUSIONS AND RECOMMENDATIONS
Conclusions
It was concluded that the DAF system performed better for Lake Chivero raw water treatment in terms of turbidity and chlorophyll-a compared to CWTPs. However, the adoption of DAF is likely to result in an increase in EC and a pH reduction. Nevertheless, it is feasible to adopt DAF for efficient algae removal and longer filter runs leading to higher output at MJWTW.
Recommendations
Further tests to be carried on determining the optimum acid dosage for DAF while using alum as a coagulant.
ACKNOWLEDGEMENTS
This paper presents part of the research results of a BSc in Civil Engineering study by Farai Mwaisowa at the University of Zimbabwe under the supervision of Eng. Z. Hoko. The authors acknowledge the financial support provided by Interconsult Zimbabwe (Pvt) Ltd for the fabrication of the bench-scale models and the assistance from the University of Zimbabwe. Finally, we appreciate the cooperation of the City of Harare in allowing sample collection at Morton Jaffray Water Works for the study.
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.