Water reuse options for decentralised MBR effluents: a case study in South Africa Open Access

Groundwater makes up about 90% of the earth's freshwater (significantly more than the surface water), which is about 1% of the total earth's water (Water Science School 2019). Groundwater utilisation is particularly popular in arid and semi-arid regions, where it is used for domestic and non-domestic purposes such as irrigation. Considering the global threat to water security and the importance of groundwater to water supply, it is necessary to ensure that groundwater sources are sufficiently protected to secure a sustainable future water supply. Human activities play a significant role in groundwater contamination due to the direct interaction between human activities that involve releasing chemicals and waste into the environment, surface water flow, and groundwater.

South Africa's water supply faces significant challenges due to its arid climate, with annual rainfall averaging only 450 mm – well below the global average of 860 mm (Hedden & Cilliers 2014). The country has a high groundwater potential, which is primarily stored in hard rock fractures and pore formations. This potential is, however, being threatened because impervious surfaces such as buildings, roads, parking lots, and pavements in continuously urbanised areas reduce surface infiltration rates as surface water from rainfall is collected and transported via pipelines or culverts and discharged into rivers, dams, or sea and is not permitted to seep into the underground aquifers (Frazer 2005). However, the country only utilises 15% of its groundwater potential, and the major reason for this underutilisation is due to the reliance of historical water management systems on surface water (Nkondo et al. 2013).

Recent trends have shown that during dry periods, people have turned to groundwater abstraction to counter the increasing pressures of drought (Arinaitwe & Okedi 2024). The increase in private boreholes (BHs) has brought about concerns regarding BH depletion due to mismanagement. Moreover, despite the refined and developed water resource management legislation in South Africa, there is a possibility that water is being extracted without appropriate regulations, resulting in potential groundwater pollution and associated health risks (Department of Water Affairs 2010). It is hence notable that continued urbanisation, population growth, and climate change would adversely impact South Africa's freshwater resources in terms of quality, quantity, and environmental pollution.

With its reliance on surface water, it has been projected that South Africa will not be able to meet its future water demands (Nkondo et al. 2013). This has resulted in a demand for utilities to seek alternative water sources to supplement their potable water resources. Reusing effluent from wastewater treatment plants is increasingly recognised as the most viable option due to its environmental benefits and cost-effectiveness when compared to seawater desalination (Turner et al. 2015). South Africa is recognised as a pioneer in potable water reuse, with established large-scale water reuse schemes such as the Atlantis Water Supply Scheme (AWSS) (Bugan et al. 2016) and several direct potable reuse (DPR) projects mentioned in Turner et al. (2015). Additionally, several large-scale water reuse programmes are currently being developed under Cape Town's New Water Programme (NWP) (City of Cape Town 2023). However, despite its pioneering status, only about 5% of the country's potential capacity for water reclamation is currently being utilised (Swartz et al. 2022). This indicates that wastewater effluents are significantly underused as a resource.

Managed aquifer recharge (MAR) – which involves artificially recharging an aquifer with stormwater or treated effluents and stormwater – is becoming a popular practice, particularly in arid and semi-arid regions (Okedi 2019). Treated wastewater offers a constant supply compared to natural fluctuations in groundwater, with no adverse health effects documented at certain levels of incorporation (Green et al. 2011). Some notable large-scale examples of MAR schemes using reclaimed water include the Dan Region Project, Israel (Icekson-Tal et al. 2003), the Gaza Strip water reuse projects, Palestine (Hamdan et al. 2011), and Atlantis, South Africa (Bugan et al. 2016). With proper design, advanced technology, and effective management, reclaimed wastewater can become a valuable resource (Asano & Cotruvo 2004).

Among the various technologies used for tertiary treatment in water reclamation projects, the membrane bioreactor (MBR) is a proven method that integrates biological treatment with a membrane separation process. This results in high-quality effluent that is low in particulate and organic matter and hence facilitates effluent reuse. The performance and operating experiences in municipal wastewater reclamation and reuse projects are widely reported in the literature. This includes experiences from projects such as the Point Loma Reclamation Project in San Diego, CA, USA (DeCarolis & Adham 2007; DeCarolis et al. 2009) and the Ulu Pandan Water Reclamation Plant (Lay et al. 2017). Besides their application in municipal wastewater treatment, MBR technology has also emerged as a suitable technology for decentralised wastewater treatment (Meuler et al. 2008). Decentralised MBR treatment plants are commonly found in wealthier rural and urban areas where water quality is an issue or in rural, informal urban, or peri-urban communities where centralised wastewater treatment systems are unavailable. Several case studies, particularly in Europe, indicate that effectively operated MBR treatment plants have a great potential for enhancing effluent reusability and reducing freshwater consumption in centralised and semi-centralised facilities (Meuler et al. 2008; Rodríguez-Hernandez et al. 2013).

In South Africa, the increasing water scarcity and frequent drought occurrences have forced communities to explore alternative options for water supply. During dry periods, individuals have turned to groundwater abstraction to counter the increasing pressures of drought. The increase in private BHs has brought about concerns regarding BH depletion due to the mismanagement by pumping them at higher rates than the rate at which the groundwater can enter the BH. To ensure sustainable water management, treated effluents could be used to recharge groundwater for later abstraction. However, using treated wastewater for groundwater recharge raises concerns about health risks associated with pathogens and chemicals. Inadequately treated wastewater effluents could contaminate groundwater sources and pose health risks to consumers downstream of the artificially recharged aquifer. Studies in the USA have shown that Giardia and Cryptosporidium are common waterborne diseases originating from groundwater, with symptoms including gastrointestinal issues (Craun et al. 2006). Chemical contamination is a lesser concern in terms of health risk than pathogenic contamination and only becomes hazardous after prolonged exposure, causing physical impairments and carcinogenic effects. The two main chemical contaminants of concern for health risk are fluoride and arsenic (Howard et al. 2006). Addressing these concerns is crucial to ensure the safety of reclaimed wastewater used for groundwater recharge. Hence, the technical and health challenges must be evaluated before the project is implemented.

This study explores the potential reuse of treated wastewater from decentralised MBR treatment plants, featuring case studies from three sites in the Western Cape province of South Africa. The specific objectives were to (i) characterise the effluents from these three sites to assess their suitability for irrigation reuse without further treatment and (ii) design and conduct a cost analysis of an advanced treatment system that effectively treats MBR effluents to meet the South African National Standard for drinking water (SANS 241 Class 1).

The three sites experience a Mediterranean climate, with winter occurring from May to September and summer from December to March. The climate graphs for Paarl and Ashton are illustrated in Figure 2.

The geology of Paarl and Ashton consists of Precambrian Basement Rock, which is superimposed by volcanic rock and marine and continental sediments characterised as Phanerozoic rocks. The Palaeozoic sedimentary, igneous, and plutonic rocks overlie the Phanerozoic rocks and the Malmesbury Group that are exposed at the surface. The Malmesbury Group is typically a softer rock building up in the valley areas (Department of Environmental Affairs and Development Planning (DEADP) 2011). The Malmesbury fractured aquifers are generally low-yielding, but in areas where fractures are well-developed, they can be high-yielding. Sites A and B are situated on a fractured aquifer that has good quality water, with yields ranging from 0.1 to 0.5 L/s (Department of Environmental Affairs and Development Planning (DEADP) 2011). The groundwater recharge in the area is considerably high (27–36 Mm³/a) due to the high precipitation rates (Department of Environmental Affairs and Development Planning (DEADP) 2011). The fractured aquifers near Site C are high-yielding and generally good quality water, ranging from 0.5 to 2 L/s (Department of Environmental Affairs and Development Planning (DEADP) 2011). The groundwater recharge in the area is generally lower, ranging from 2 to 3.5 Mm³/a (Department of Environmental Affairs and Development Planning (DEADP) 2011).

The decentralised MBR plants were designed to treat raw sewage from various regions within Sites A, B, and C. The decentralised MBR treatment plants consist of a coarse screen, buffer tank, lifting station, fine screen, anoxic process, aeration process, MBR filtration, return activated sludge, and disinfection. Design parameters for each decentralised MBR treatment system are provided in Table 1.

Groundwater data, such as geological, hydrological, water quality, and recharge estimates, were obtained from the existing literature, primarily from various hydrogeological reports in the DWS's geographic information system databases. Since there are no existing BHs in the selected sites, data from the surrounding locations were utilised for this exercise. Sites A and B used the same data, which were obtained from BHs located in Val de Vie (Geohydrological and Spatial Solutions (GEOSS) International 2015) and Boy Louw (GEOSS 2018a). Data from 15 BHs were considered in total, including six in Val de Vie, three in Boy Louw, and six in Montagu. The Val de Vie and Boy Louw BHs are approximately 5–8.5 and 7–15 km away from Sites A and B, respectively. Site C utilised data from the data analysis for Montagu BHs (GEOSS 2018b), which are located approximately 15–20 km away from Site C.

The MBR effluent quality was obtained through field and experimental investigations whereby 2-hourly composite samples were collected from each plant and tested for various parameters. Samples were collected from all three MBR plants and were tested for various parameters depending on the specific water reuse water quality requirement. Sample testing was conducted by South African National Accredited System (SANAS)-accredited facilities. For statistical significance, three tests were conducted, and for each sample, the average was used as a representative.

For this study, it was assumed that the treated wastewater is being spread on the surface or irrigated to infiltrate the aquifer. The effluent from each site was tested for pH, electrical conductivity, chemical oxygen demand (COD), and faecal coliforms and analysed for irrigation water requirements in accordance with the National Water Act (Ministry of Water and Environmental Affairs: South Africa 2013). Recommendations were made for any parameters that fell out of the irrigation standards.

The indirect potable reuse (IPR) scenario replicated a process whereby the MBR effluent is directly injected into the injection wells, abstracted later, and treated in a multi-stage filtration water treatment plant to meet drinking water quality standards. The design of the MAR unit was based on the characteristics of the effluent of the MBR wastewater treatment plant (WWTP). The effluent samples collected from each plant underwent a comprehensive analysis of various parameters in accordance with the South African National Standard (SANS 241) for drinking water quality (South African Bureau of Standards (SABS) 2005). Moreover, the recharging effluent rate was adopted from a study by Green et al. (2011), which recommended that the effluent recharge volume be 25% of the wastewater treatment's demand to prevent adverse health effects on the supplied population. The results of this simulation were analysed to determine changes in groundwater volume and quality. Based on these simulation results, an advanced treatment was designed to correct the parameters that can potentially lead to contaminating groundwater if MBR effluent is recharged into the ground without additional treatment.

A multi-stage filtration package water treatment was adopted for treating water from an aquifer that is partially recharged with MBR effluent. A typical multi-stage filtration plant consists of several pressure filters containing various filter media and may include pH corrections, chemical stabilisation, ozone treatment, microfiltration, ion exchange, adsorption, and disinfection.

The DPR system replicated a scenario where MBR effluent was directly treated to drinking water standards in a reverse osmosis (RO) water treatment plant. The design of the RO plant was based on the characteristics of the MBR plant, which were also used to design the MAR unit of operation for the IPR system discussed in the previous section. For each site, a RO plant was designed to treat the MBR effluent of the respective site, such that the Class I drinking water standard (SANS 241) (South African Bureau of Standards (SABS) 2005) is met. The RO plant was designed according to the WAVE software (DuPont Water Solutions n.d.; Roopchund et al. 2022). Prior to designing the RO, the need for MBR effluent pre-treatment was evaluated. A pre-treatment unit helps prevent fouling and scaling of membrane products, thereby optimising membrane performance and extending its lifespan. Final disinfection to eliminate pathogens and coliforms from the treated water was achieved by using UV lights. Additionally, pH correction was accomplished by dosing sodium hydroxide. The design of the pre-treatment unit, disinfection unit, dosing rates, and pH correction followed the same approach applied in the design of the multi-filtration system (see Appendix A).

It was assumed that sewage produced equated to 95% of the water demand; a discounted rate of 12% applies. Tariffs were based on the 2019/2020 City of Cape Town Budget (City of Cape Town 2019).

Table 2 shows the MBR effluent water quality analysis and the requirements for irrigation water quality. Generally, all locations passed the irrigation standards except for Site C, which will require a pH adjustment to meet the irrigation standard. Typical alternatives for pH adjustment include a system consisting of an appropriately sized dosing pump, chemical makeup tank, sodium hydroxide chemical drums, and a control panel. Additionally, exposing the effluent to a calcium carbonate filter or bank can help stabilise the water (Meenakshipriya et al. 2008). The same MBR effluent results were used to evaluate whether MBR effluents are suitable for potable reuse, as discussed in the next section.

Table 2

View Large

The water quality results for MBR effluent and BH groundwater are summarised in Table 3 alongside Class I drinking water requirements. The MBR effluents from both locations yielded poor drinking water quality, with 25–35% of their parameters falling outside the Class I drinking water quality standard (acceptable water quality). In particular, these effluents required treatment primarily to improve mainly pH, Langlier Saturated Index (LSI), colour, turbidity, fluoride, nitrates, ammonia, total organic carbon (TOC), and COD.

Table 3

View Large

On the other hand, groundwater quality analysis from 14 out of 15 BHs generally yielded Class 0 water quality (ideal for domestic water quality), with only iron, manganese, and faecal coliforms identified as parameters requiring treatment. Only one BH (in location C), yielded poor water quality, with 56% of its parameters falling within the Class III category. Water in this category is considered economically not feasible to treat to the desired Class I drinking water quality. However, this BH was excluded from further analysis in this study.

The results of a blend of groundwater and MBR effluent, which represent the quality of groundwater partially recharged with MBR effluent, are also presented in Table 3. The results indicate that recharging groundwater with MBR effluent did not lead to a significant improvement in water quality. Generally, the results indicate that recharging MBR effluent without additional treatment can contaminate groundwater further beyond drinking water quality, particularly in terms of organic matter, faecal coliforms, and nitrate parameters. Despite the low turbidity and concentrations of iron and manganese in the MBR effluent, recharging 25% of this effluent is insufficient to improve groundwater quality parameters to meet drinking water standards. As a result, partially recharged groundwater for all three sites requires additional treatment in order to meet Class I drinking water quality. Generally, the parameters that require treatment are the LSI, colour, turbidity, iron, manganese, faecal coliforms, nitrates, and organics (TOC and COD). For this study, the multi-stage filtration water treatment system was designed to target the treatment of these parameters within the Class I drinking water requirement.

The multi-stage filtration treatment was designed based on the worst-case scenario whereby the highest BH concentration of each parameter was considered when implementing 25% of the treated effluent to form a partially recharged aquifer. For each location, the plant was designed to target the treatment of the parameters that do not fall within the Class I drinking water requirement. The design of the treatment system did not take into account the removal of colour because this is usually removed through flocculation/coagulation, and other processes such as organics and metals (iron and manganese) also aid with colour removal (Schutte 2007).

LSI, which is an indicator of how the water is saturated with CaCO₃, shows that partially recharged groundwater is undersaturated. Water that is undersaturated with CaCO₃ is corrosive and requires pH stabilisation. This can be achieved by introducing limestone media, a naturally occurring calcium carbonate. Limestone filters for the multi-stage filtration process were sized according to Lenntech (n.d.) and Pure Water Products (n.d.). The dimensions of the limestone filter media at all three locations were nearly identical, with diameters ranging from 0.5 to 0.7 m and heights ranging from 0.5 to 0.9 m. On the other hand, the turbidity of the partially recharged aquifer was higher than the required class water quality, which suggests that the water is murky. The Activated Filter Media (AFM) was selected to reduce water's turbidity and total suspended solids (TSS) in the multi-stage filtration design. The AFM filters for all three locations were identical, with diameters and heights of 0.5 and 1 m, respectively.

Iron and manganese were removed using a combination of pH control, ozone treatment, and media filtration. The pH of the feed water was maintained between 8 and 8.5 with sodium hydroxide, while limestone also helped to raise the pH, allowing the BIRM media to activate effectively. Ozone was used to oxidise the iron and manganese, providing the necessary 15% dissolved oxygen (DO) required by the BIRM. Ozone and granular activated carbon (GAC) were designed for the removal of excess TOC, COD, and nitrates. Ozone was also designed as an additional treatment barrier for pathogen disinfection (pathogens were not analysed in this study). Hence, a high dosage of 1.5 mgO₃/L required for cryptosporidium disinfection was selected. The diameter of BIRM and GAC media typically ranges from 0.5 to 0.7 m, while their height is 0.5 m for all three sites. Last, microfilters were added to remove excess suspended solids after the series of filter media and final disinfection through ultraviolet (UV) lights.

Table 4 provides a summary of the multi-stage filtration treatment plant equipment required to treat partially recharged groundwater for each of Sites A, B, and C.

An RO water treatment plant is typically used to treat wastewater to potable water standards. The design of the RO process was based on the MBR effluent data, which is presented in Table 3. The concentration of several parameters of the MBR effluent, namely turbidity, did not meet the feed water specification for the RO treatment plant, which necessitated the need for a pre-treatment unit. A pre-treatment unit helps prevent fouling and scaling of membrane products, optimising membrane performance and extending its life span. The pre-treatment unit consisted of AFM and GAC for reducing turbidity, TSS, COD, and TOC. The AFM and GAC were sized using a similar approach that was used to size the AFM and GAC for the multi-filtration treatment system discussed above. The removal of colour was excluded from the RO treatment design because it is usually removed through the coagulation/flocculation process and can also be further reduced indirectly through the removal of organics and metals. A design summary of the RO treatment system for each site is presented in Table 5.

The capital expenditure (Capex) of the IPR scheme considers the drilling and casing of the BHs, geohydrological study fee, BH pumps and pipe fittings, water treatment plant mechanical and electrical equipment, and the cost of civil works (i.e., excavation, trenching, concrete platform, and steel canopy). The Opex takes into account the operation of the multi-stage water treatment plant and considers all the costs required for the treatment plant to operate successfully on a day-to-day basis. This includes media, chemicals, equipment, and labour. Table 6 provides the summary of Capex and Opex costing for each site.

Across all three sites, BHs, water treatment systems, and civil work accounted for 20–22, 51–54, and 26–27% of the total Capex costs, respectively. Treating groundwater that is partially recharged with treated effluents is an intricate process that requires an automated and remote monitoring system, which makes it more expensive and, thus, the high cost of the water treatment system. While the Opex of the water treatment system rises with increased flow rates, this increase is not linear. Consequently, the analysis reveals that the cost per cubic metre goes up as the treatment plant's flow rate decreases.

Table 7 summarises the Capex and Opex costs based on a 31-day month. For the IPR system, the Capex of the RO treatment process considered the mechanical equipment, electrical equipment, and the cost of civil works. The study determined that RO treatment represents about 70% of the total Capex. The RO treatment plant's higher cost is justified because it was designed on a recovery of 60% to make it more cost-effective. Additionally, the monitoring system is designed to protect RO membranes and minimise contamination risks in drinking water due to its DPR operation, which significantly increases RO costs. The monitoring system includes automated and remote controls for quick and easy troubleshooting when faults occur. The Opex of the RO water treatment plant considered all the costs required for the treatment plant to operate successfully on a day-to-day basis, including filter media, chemicals, electricity, equipment, and labour.

This paper evaluated wastewater reclamation and reuse strategies from three decentralised MBR plants in Western Cape, South Africa. The assessed reuse strategies included non-potable reuse for irrigation, IPR via MAR, and DPR. Additionally, the study calculated the payback period for investments in the two potable reuse options for each MBR plant. The primary aim was to assess feasible wastewater reuse options from MBR plants with different design and operational conditions. The analysis of raw MBR effluent from all three sites reveals that these effluents are suitable for irrigation purposes with minimal additional treatment needed.

The analysis of the potable reuse option indicates that although both IPR and DPR are technically feasible, only one site is economically viable for implementation. The payback periods are 8 years for the IPR system and 7 years for the DPR system. Further, the results show a trend whereby return on investment reduces as the flow rate decreases, which implies that, although return on investment is case-specific, generally, investment in potable reuse projects is feasible for large MBR plants and economically unviable for small-scale MBR plants. Therefore, for economic reasons, unless the necessity for potable reuse during severe drought outweighs the financial implications to become viable, potable reuse is not recommended for small-scale MBR plants (i.e., Sites B and C). Although non-potable water reuse is technically feasible for small-scale MBR plants and is expected to be economically viable, conducting an economic evaluation may be necessary to ascertain the true economic value of the investment.

Future research should focus on the following:

• Identifying and evaluating cost-effective treatment technologies for the advanced treatment of MBR effluents to meet potable water standards.

• Integrated modelling of the MAR scheme to assess the groundwater and surface water interaction along with the recharging MBR effluent and to evaluate the potential for various MAR scenarios.

• Extending the investigation to incorporate the evaluation of the occurrence and fate of micropollutants, including contaminants of emerging concern in such reuse schemes.

The authors thank Alveo Water (Pty) Ltd for funding this project through a master’s degree funding for Enaas Richards and for their support during the sample collection process.

E.R. undertook the research including the literature review, the selection of a suitable method, the development of the assessment model, the generation of the results, wrote the draft paper, and implemented the revisions from feedback/comments received from the co-authors and reviewers. S.T.A. (https://orcid.org/0000-0002-1152-6601) assisted in writing, reviewing, and editing. W.B.A. (https://orcid.org/0000-0002-2930-9728) assisted in writing, reviewing, and editing the paper for submission. J.O. (https://orcid.org/0000-0001-7707-2721) assisted in supervision, paper writing, formatted the paper for submission, and is the corresponding author. D.S.I. proposed the research topic, managed and supervised the study, assisted in the paper writing, formatted the paper for submission, and is the corresponding author.

Alveo Water provided funding for this research project.

No human or animal participants were involved in this study

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.