Application of a basic monitoring strategy for Cryptosporidium and Giardia in drinking water

Despite the health risks associated with exposure to Cryptosporidium and Giardia, there is no uniform approach to monitoring these protozoan parasites across the world. In the present study, a strategy for monitoring Cryptosporidium and Giardia in drinking water was developed in an effort to ensure that the risk of exposure to these organisms and the risks of non-compliance to guidelines are reduced. The methodology developed will be applicable to all water supply systems irrespective of size and complexity of the purification works. It is based on monitoring procedures proposed by the US Environmental Protection Agency, the Drinking Water Inspectorate, Australia and New Zealand, as well as the risk-based procedure followed by Northern Ireland. The monitoring strategy developed represents a preventative approach for proactively monitoring Cryptosporidium and Giardia species in drinking water. The strategy consists of 10 steps: (i) assessment of the monitoring requirements, (ii) description and characterisation of the source water types, (iii) abstraction of source water, (iv) assessment of the water purification plant, (v) water quality monitoring, (vi) cryptosporidiosis and giardiasis outbreak, (vii) risk assessment, (viii) sample collection and laboratory processing, (ix) data evaluation, interpretation and storage, (x) process evaluation and review. Proper implementation of this protocol can contribute to the protection of drinking water consumers by identifying high-risk source water, identifying areas of improvement within the water treatment system, and also preventing further faecal pollution in the catchments. The protocol can also be integrated into the Water Safety Plans to optimise compliance. Furthermore, this methodology has a potential to contribute to Blue Drop certification as it should form part of the incident management protocols which are a requirement of Water Safety Plan implementation.


INTRODUCTION
Cryptosporidium and Giardia are intracellular protozoan parasites that infect the gastrointestinal tract of vertebrate animals including mammals, birds, reptiles, and fish (Carmena, 2010). They enter surface waters such as lakes, ponds and dams as environmentally resistant cysts and oocysts in the faeces of infected people or animals. Cryptosporidium and Giardia can cause human cryptosporidiosis and giardiasis, respectively, which are the most common causes of protozoal diarrhoea worldwide (Cacciò et al., 2005). The transmission of Cryptosporidium and Giardia occurs mainly through the faecal-oral route by direct contact with contaminated faeces and by exposure to contaminated food and water (Carmena et al., 2012). The water exposure routes include ingestion of water during recreational activities (swimming, canoeing, and skiing), and drinking untreated and treated tap water contaminated with Cryptosporidium and Giardia oocysts.
Monitoring of Cryptosporidium and Giardia is done throughout the world and the data obtained have been used for conducting risk assessment, for evaluation of water treatment system reliability and also to assist with waterborne outbreak investigations (Bentacourt and Rose, 2004). However, different approaches for Cryptosporidium and Giardia monitoring are used in different countries (US EPA, 2006;DWI, 2008;NZ Ministry of Health, 2008; NI Department for Regional Development, 2007). In the USA the log-reduction method is used (US EPA, 2006), while risk-based approaches are commonly applied in the United Kingdom and New Zealand (DWI, 2008;NZ Ministry of Health, 2008). In South Africa, Cryptosporidium and Giardia monitoring in drinking water supply systems is not pervasive. In the few systems where it is performed, the focus is mainly on the final treated water.
From the preceding, it is evident that there is no uniform strategy for monitoring these protozoan parasites in drinking water supply systems. Hence the overall objective of the study was to develop and apply a methodology that can be used by drinking water utilities to monitor Cryptosporidium and Giardia. The test site used was a small potable water treatment plant using surface water abstracted from the Vaal River Barrage, Gauteng Province as source water. It must be stressed that this methodology is based on monitoring procedures proposed by the US Environmental Protection Agency (USEPA) (2006), the Drinking Water Inspectorate (DWI, 2008), Australian Drinking Water Quality Guidelines (NHMRC, 2004), Drinking Water Standards New Zealand (NZ Ministry of , and especially on the riskbased procedure followed by Water Supply Water Quality Regulations Northern Ireland (NI Department for Regional Development, 2007

Elements of the protocol
This protocol involves desktop studies, surveys, as well as laboratory analysis with scores assigned during the surveys. It includes 10 major steps that should be followed sequentially for monitoring of Cryptosporidium and Giardia by water utilities (Fig. 1).
These steps are grouped into 4 major phases, namely: At the end of each assessment the total score can be determined. These are then used to calculate the risk of Crypto sporidium/Giardia contamination.

PHASE I: Desktop survey of monitoring requirements
This phase involves the accessing of current information on 4

STEP I: ASSESSMENT OF THE MONITORING REQUIREMENTS
 What is the scope and the level of the assessment?  What are the objectives of the survey?

Literature review
Evaluate the current status of Cryptosporidium and Giardia monitoring

Objectives of the survey?
 National legal requirements  International requirements

PHASE I: DESKTOP SURVEY OF MONITORING REQUIREMENTS STEP IV: ASSESSMENT OF THE WATER PURIFICATION PLANT
 What is the type of water treatment system?  Are the different steps of water treatment monitored?
 Is filter performance evaluated?  Is there additional treatment?
 What quality assurances are followed?
Step II (a): Types of source water  Surface water  Ground water  Borehole

STEP II: DESCRIPTION AND CHARACTERISATION OF THE SOURCE WATER TYPES AND ACTIVITIES AROUND THE CATCHMENT AND SOURCE WATER
 What are the types of source water?  Identify activities around the catchment and source water

High Risk Area
Risk score 50 -99

Medium Risk Area
Risk score 10 -49

Low Risk Area
Risk score < 10

STEP VI: CRYPTOSPORIDIOSIS AND GIARDIASIS OUTBREAK
Are there any known cases of Cryptosporidium and/ Giardia infection? Is there Cryptosporidium and/or Giardia awareness?

Very High Risk Area
Risk score> 100 299 Cryptosporidium and Giardia from published and non-published sources and forms part of the first step of the protocol. However, if improvements in technology have been made or when there are changes in the national standards and acts, it is advisable that the water utility revisit the sites to update the protocol.
Step I: Assessment for monitoring requirements (Fig. 2) This is a desktop study for evaluating monitoring strategies required for Cryptosporidium and Giardia. During this step, the national water related legislation (Acts and standards) must be scanned in order to understand the legal requirements for Cryptosporidium and Giardia in drinking water and source water. This will give an insight into current minimum requirements for monitoring Cryptosporidium and Giardia in water bodies used for domestic purposes, and agricultural as well as recreational purposes. In South Africa, national acts and standards apply, such as the South African National Standard (SANS) 241 (2011) for drinking water, National Water Act (NWA) (No. 36 of 1998) (RSA, 1998), and Water Services Act (WSA) (No. 108 of 1997) (RSA, 1997).
The international legal requirements should also be reviewed to gain further insight into the current global practices regarding Cryptosporidium and Giardia monitoring. The international standards and acts to be scanned include those of the U.S. Environmental Protection Agency (US EPA, 2006), Drinking Water Standards New Zealand (NZ Department of , Australian Drinking Water Quality Guidelines (NHMRC, 2004), the Drinking Water Inspectorate (DWI, 2008), the Water Supply (Water Quality) Regulations Northern Ireland (NI Department for Regional Development, 2007), and other relevant international acts and standards applicable during the assessment.

Application
The site used to test the protocol was the Vaal River Barrage Drinking Water Production Plant which uses the water from the Vaal River Barrage Reservoir as source water. The Vaal River Barrage Reservoir is situated on the lower Vaal River in Gauteng (South Africa) and covers an area of approximately 64 km 2 with an estimated storage capacity of 63 x 10 6 ℓ (RW, 2008).
Two main approaches for monitoring protozoan parasites have been identified. These are the log-reduction approach, applied in the United States (USEPA, 2006) and New Zealand (NZ Department of , and the risk-based assessment followed in the United Kingdom (DWI, 2008) and Northern Ireland (NI EHS, 2002). In Cryptosporidium and Giardia monitoring in other countries such as Canada there are no minimum acceptable concentrations for these protozoans (Health Canada, 2008).
In South Africa drinking water quality is evaluated against the South African National Standard (SANS) 241-1. According to SANS 241-1 (2011) the operational water quality values for Cryptosporidium and Giardia in drinking water are less than 1 oocyst/10 ℓ and 1 cyst/10 ℓ, respectively. The standard further stipulates that the final water be monitored monthly for Cryptosporidium and Giardia. Yet, not all water utilities in South Africa monitor for these parasitic protozoans as there is no standard protocol for their monitoring. The drinking water produced by the Vaal River Barrage Drinking Water Production Plant must therefore comply with SANS 241-1 (2011), and Rand Water monitors the levels of Cryptosporidium and Giardia in the final and source water of the plant every second week.

Phase II: Situation analysis of source, water purification plant and disease outbreak
This phase involves conducting an on-site assessment of the source water, facilities, equipment, operation and maintenance of the entire drinking water system. This is an intensive phase of the survey which includes Steps II -VII of the protocol. During this survey the source water and the water treatment plant can be evaluated for the risk of Cryptosporidium and Giardia contamination. At the end of this phase the risk of the complete water treatment system can be categorised into either 'very high risk', 'high risk', 'medium risk' or 'low risk'.
Step II (a): Description and characterisation of the source water types (Fig. 3) The source water is characterised in order to identify the possibility of Cryptosporidium and Giardia contamination. Source water types could be classified into: • Protected groundwater -groundwater is usually protected against contamination from the surface by soils and covering rock layers (ESSF, 2005). In other cases groundwater may be protected by defining source protection zones within which development is limited in order to reduce the chance of spillage of potentially polluting substances • Unprotected groundwater -is not protected from possible impact from human sewage or livestock faecal material (NZ Department of . • Protected surface water catchments -as with groundwater, 300 surface water could be protected by defining source protection zones within which development is limited in order to reduce the chance of spillage of potentially polluting substances.
• Protected surface water usually contains low to moderate numbers of E. coli and should be considered as being vulnerable to contamination by enteric viruses, protozoa and bacteria (NZ Department of . • Unprotected surface water -lacks defined protection zones and can be subject to contamination by enteric viruses, protozoa and bacteria (NZ Department of . • Reclaimed water -reclaimed water for potable use is not protected from faecal contamination.
The monitoring requirements for water supplies derived from protected groundwater are different from those for water supplies derived from unprotected surface water. Protected surface water catchments should contain low-moderate numbers of E. coli, whereas unprotected surface water can be subject to contamination by enteric viruses, protozoa and bacteria (US EPA, 2006). Any observations of the source water that could indicate the possibility of faecal contamination become of importance as it may assist in risk classification given that human sewage and livestock faecal material are the predominant source of human-infective Cryptosporidium and Giardia (NZ Department of .

Application
At the Vaal River Barrage plant, unprotected surface water is used as source water. Visual assessment at the Vaal River Barrage Drinking Water Production Plant abstraction point indicated that there was no evidence of faecal pollution or offensive odours. The bottom sediment was muddy and the abstraction is done at the reservoir wall (Fig. 3).
Step II (b): Assessment of the activities around the catchment and the source water ( Fig. 4) The activities around the catchment and source water should be assessed in order to determine the processes that drive and determine the characteristics of the water quality in that specific source water and the catchment (Du Preez et al., 2003). The quality of drinking water sources is threatened by land development, runoff from agricultural, commercial and industrial sites, farmed animals, ageing wastewater infrastructure, and urbanisation, as well as the type of settlement around the source (Murray et al., 2004). Since certain land use activities may pose a high risk, due to causing faecal pollution, an initial land-use survey of the catchment should be undertaken to: identify existing and planned developments, potential continuous, intermittent or seasonal pollution patterns and geological features likely to affect water quality and any existing or abandoned waste-disposal sites (NHMRC, 2004).

Application
The activities around the catchment and the source water were assessed following the steps in Fig. 4 (2003) and Osche (2007). The increase in urban population, informal settlements, wastewater treatment plants and industrial complexes in the Vaal River Barrage catchment may have an impact on the rivers discharging into the Vaal River Barrage. A third of the Rietspruit catchment area is covered by informal settlements, urban areas, wastewater treatment plants, industry and mining (Van Baalen, 2003). Some of the formal and informal settlements in the catchments of the rivers feeding into the Vaal barrage do not have proper sanitation facilities.
Septic tanks and stormwater outlets are also present, increasing the risk of discharging faecally-contaminated water into the source water. The discharge of partially-treated wastewater may result in high concentrations of faecal micro-organisms of 302 concern, such as E. coli, Klebsiella and parasitic protozoans such as Cryptosporidium and Giardia in the source water (Pitt, 2007). This is confirmed by Dungeni and Momba (2010), who reported that up to 400 Cryptosporidium oocysts/10 ℓ and up to 1 750 Giardia cysts/10 ℓ were detected in the effluents of some of the wastewater treatment plants in Gauteng. Several events that could lead to increased faecal pollution on the Vaal River catchment have been reported. These include pumping of raw sewage into the Vaal River Barrage (FSE, 2009) and frequent overloading of the sewage works as a result of heavy rains, which results in spillage of the effluent into the Vaal River Barrage catchment (MG, 2006). These events could pose a serious health risk since untreated sewage samples have been found to contain an average of 1 000 -44 5000 Cryptosporidium oocysts/10 ℓ and 1 000 -51 3330 Giardia cysts/10 ℓ (Robertson et al., 2006).
Source water monitoring activities also revealed that some of the rivers, such as the Rietspruit, Klip River and Suikerbosrand River ( The presence of sheep farming, cattle farming and piggeries was observed at households in the catchment. The presence of even small farmed animals in the catchment could pose a risk since infected lambs, calves and goat kids have been reported to shed a significant number of cysts and oocysts between the ages of 5 and 25 days (Sari et al., 2009). Additionally, the piggeries (at Groenpunt Prison) discharging waste near the inlet of the Vaal river Barrage reservoir also pose a risk. Intensive stock farming (Karen Beef) was observed, which included an abattoir producing wastewater draining into the Suikerbosrand River.
Step III: Abstraction of the source water (Fig. 6) This step allows for evaluation of the risk associated with the quality of water at the abstraction point. The quality of the water at the intake determines the ability of the treatment plant to effectively treat the water, which could pose a health risk. Monitoring at the abstraction point will assist in establishing the levels of source water contamination that can be handled. Monitoring at this point can also identify mechanical failures during abstraction as well as incidences such as power failures, which can assist in avoiding sourcing too little water from the intake to meet the demand (NZ Ministry of Health, 2001).

Application
During the assessment it was observed that the abstraction point at the Vaal Barrage catchment lacked online water quality  Step IV: Assessment of the water treatment plant (Fig. 7) It is critical to monitor the performance of the water treatment plant since it plays a significant role in ensuring that water from the catchment is treated in order for it to be safe for consumers to drink. For the water treatment plant to be completely effective in the removal of Cryptosporidium/Giardia from the source water the plant should be operating optimally with no interruptions.
Monitoring of plant performance includes assessing the situations in the plant that could lead to the breakthrough of Cryptosporidium and Giardia oocysts (NHMRC, 2004). The following areas of the water treatment plant are to be evaluated: • The type of water treatment processes. Processes that should be evaluated include coagulation/flocculation, sedimentation, carbonation, filtration and disinfection (multibarrier approach). This approach has been reported to be effective for the removal of Cryptosporidium and Giardia oocysts. The use of disinfection only as well as the use of additional treatment (i.e. ozone treatment or UV treatment) will be evaluated (WHO, 2006) • Operational parameters. Assessment of monitoring in the treatment plant will include monitoring of turbidity after coagulation, monitoring of residual coagulants, monitoring of floc settling rate, inspection of water clarity after sedimentation, as well as evaluation of pH during carbonation. • Filter performance and turbidity during filtration. It is important to ensure that there are no disturbances in the flow and that the filter nozzles are in a good condition since this could result in poor particle removal (Haarolf, 2008). Monitoring of the turbidity resulting from each of the individual filters should be conducted to ensure that filtration is effective (NHMRC, 2004). For effective removal of Cryptosporidium oocysts and Giardia cysts it is important to maintain the turbidity at less than 0.5 NTU (NZ Department of Health, 2008). • Quality control. This step will assist in evaluating the presence of a quality control system and the level of adherence thereto. For the treatment works to operate optimally it is important to ensure that all deviations to the quality control procedures are attended to. An example of the quality control systems that could be put in place for effective operation of the water purification plant are the ISO 9000 and ISO 14001.

Application
The Vaal River Barrage water purification plant uses coagulation/flocculation, sedimentation, carbonation, sand filtration, GAC treatment, UV irradiation and chlorination to treat source water (Fig. 7). A multi-barrier approach is used to achieve the water quality target in accordance with SANS 241-1 (2011). This approach is widely accepted for managing water quality and recommends the provision of water quality protection mechanisms at multiple points from the catchment to the tap (WHO, 2011). In addition a UV plant has been installed which in other studies has been reported to achieve up to 99% protozoan removal (WHO, 2006). Water flows through the UV plant where light intensity is maintained above 80%; should it fall below this level there is an audible alarm. The water is tested on a grab sample basis. Should deterioration in quality of the UV output be noted by the operator, UV light tubes are replaced (RW, 2009a). Quality should meet the SANS Class 1 drinking water standard. A chlorination step follows UV treatment. While assessing the operational parameters, it was observed that residual coagulant was not monitored, but other operational parameters such as floc settling rate, sedimentation and carbonation were closely monitored (Fig. 8). The filters were in good condition as there was no evidence of mud balls and algae (Fig. 9). No increase in the turbidity of the treated water was observed. Compliance levels of 99% were recorded during the evaluation period since it was less than 0.5 NTU most of the time. During the period January 2007 to January 2013 the average NTU was 0.32 (Fig. 10).
Assessment of the quality control system indicated that a quality system was in place (the water treatment plant is ISO 9000 accredited); hence operational manuals are available, equipment is verified, operators are trained and deviations to quality are corrected by implementing the corrective actions (Fig. 11).
Step V: Water quality monitoring (Fig. 12) In this step the availability of monitoring data for Crypto sporidium and Giardia is assessed. The availability of 304 monitoring data will give an indication of the status of the source water in terms of risk. Hence it will assist the water utility in taking the necessary precautionary measures such as selecting the treatment method to be used, and developing an applicable sampling programme before using the source water. The data for treated water can also assist the water utility in evaluating whether the treatment process in place is effective or not. This can assist in identifying if there is a need to improve the treatment process or the monitoring processes to improve the mitigation of Cryptosporidium and Giardia (NHMRC, 2004). In South Africa the thresholds for Cryptosporidium and Giardia are less than 1 oocyst per 10 ℓ; in countries such as the UK, USA and Northern Ireland the minimum acceptable concentration for Cryptosporidium and Giardia is also less than 1 oocyst per 10 ℓ.

Application
It was found that both source water quality at the point of abstraction and treated water quality are monitored for Cryptosporidium and Giardia. Source water is monitored every second week whereas treated water is monitored every week.
For the purpose of this study, Cryptosporidium and Giardia data collected from January 2007 up to January 2013 were used and the following sampling points were used (Fig. 5): • C-RV2: Sample point at the Rietspruit weir • B-Raw: Source water at the point of abstraction • B-Dom: Drinking water supply from the BDWPP • Klip River • Suikerbosrand River The data indicated that water in Vaal River Barrage Reservoir contains Cryptosporidium and Giardia cysts. Source water monitoring results revealed that some of the rivers, such as the Rietspruit, Klip River and Suikerbosrand River, feeding into the Vaal River Barrage Reservoir and at the Vaal River Barrage intake (Figs 13 -16) also contain high concentrations of Cryptosporidium and Giardia. For example, for the period January 2007 to January 2013 the average numbers of Cryptosporidium and Giardia in the Rietspruit (Table 1) were 295 oocysts/10 ℓ and 153 cysts/10 ℓ respectively, in the Klip River were 31 oocysts/10 ℓ and 50 cysts/10 ℓ, respectively, and in the Suikerbosrand River were2 oocysts/10 ℓ and 1 cyst/10 ℓ respectively. A significant increase in Cryptosporidium concentrations was observed during February2010 and June 2010. This could most likely be as a result of wastewater discharges into the Rietspruit.
Although data obtained showed the presence of   305 Crypto sporidium and Giardia cysts in the source water and the abstraction point (Fig. 16), only one Cryptosporidium oocyst was detected in the treated water and no Giardia cysts were detected in the treated water for the period January 2007 -January 2013 (Table 1). This could be attributed to the multibarrier treatment system at the Vaal River Barrage drinking   307 water purification plant as well as the presence of the advanced treatment system (the UV plant installed).

Step VI: Cryptosporidiosis and giardiasis outbreak
This step evaluates if there are any known incidences of cryptosporidiosis and/or giardiasis, whether cryptosporidiosis and/or giardiasis has been suspected and also checks awareness of the disease.

Application
The awareness of cryptosporidiosis and giardiasis was evaluated (Fig. 17), and the history of cryptosporidiosis and giardiasis from the same supply was evaluated. Results obtained

Awareness of Cryptosporidiosis and Giradiasis
Total Score -1*

Figure 17
Cryptosporidiosis and giardiasis outbreak (adapted from NI EHS, 2002) indicated that there was lack of Cryptosporidium and Giardia awareness among the community. Therefore, no cryptosporidiosis and giardiasis has been reported at the Vaal River Barrage.
In general there is a lack of cryptosporidiosis and giardiasis epidemiological data in South Africa.
Step VII: Risk categorisation (Fig. 18) In this step all the scores assigned in the previous steps are processed in order to allocate the final risk score and rating. The proposed protocol will allow for assessment of a possible risk score from the source water as well as the possible risk of Cryptosporidium and Giardia that could arise from the water treatment plant. Calculation of the final risk score is done according to 308 the formula described in the guidelines for monitoring of Cryptosporidium in Northern Ireland (NI EHS, 2002) which is as follows: Risk is classified into 4 categories which are: very high risk (risk score > 100; high risk (50 < score < 100); medium risk (10 < score < 50) and low risk (score < 10) (NI EHS, 2002).

Application
The risk score of the source water was calculated as follows: Based on the 2007 census figures which gave the population of the Vaal Triangle as 800 819 (Statistics South Africa, 2007) and a 2% annual population growth rate (Emfuleni, 2008), the population of the Vaal Triangle in 2012 was estimated to be 880 901.

For this assessment:
Population impacting the source water = 880 901 Population weighting = 0.4 log 10 (880 901) = 2.778 Final risk score for source water: = sum of total score (source water) x population weighting = (22 + 42) x 2.778 = 177.79 The risk score for the source water from the Vaal River Barrage Reservoir is 177.79, which implies that the Vaal River Barrage catchment can be regarded as a very high risk area for Crypto sporidium and Giardia. The risk score for the treated water was calculated as follows: Risk score for the treated drinking water from the Vaal River Barrage Drinking water purification plant is −24.4. This implies that the drinking water from Vaal River Barrage drinking water purification plant can be regarded as low risk with regards to Cryptosporidium and Giardia since 24.41 is less than 10. This could be attributed to the multi-barrier water treatment system as well as the UV irradiation plant operational at the Vaal River Barrage. UV irradiation has been reported to be effective for Cryptosporidium and Giardia removal where removal rates of up to 99.9% Cryptosporidium and Giardia oocysts can be achieved (WHO, 2006).

PHASE III: Sampling, analysis and review
This phase of the protocol includes sample collection, analysis, data storage and review. This phase clearly describes what to monitor, where to monitor, when and how to monitor for Cryptosporidium and Giardia. At the end of this phase it will be possible to quantify the magnitude of Cryptosporidium and/or Giardia contamination from the water sources.
Step VIII: Sampling and laboratory processing

Sampling
Sample collection is guided by risk classification since sample frequency, sampling points and the number of samples to be collected is determined by the risk score (Fig. 19). The recommendations for sampling and monitoring are given in Table 2; however, in the incidences where Cryptosporidium is detected in the final water, sampling must be done in accordance with the Incident Management Protocol described by DWAF (2007), irrespective of the risk score, until there are no oocysts detected in the treated water. The DWAF Incident Management Protocol involves increasing the sampling frequency, and notification of the municipality, Department of Water Affairs and the Department of Health.
Although the risk determines the sampling point and the sampling frequency it is important that, before sampling, the sampling programme and requirements and the protocol are discussed with the laboratory that will be in charge of the analysis. A 10 ℓ plastic carboy or container is recommended for sampling. To avoid cross contamination it is recommended that carboys used for collection of source water be dedicated to source water only and those for drinking water be specifically used for collection of drinking water only. For sampling of the surface water it is important to ensure that sediment is not introduced into the sample and this can be achieved by sampling using a 5 ℓ beaker until the required volume has been sampled (RW, 2009b).

Laboratory processing
The selected laboratory must use internationally accepted methods for the analysis of Cryptosporidium and Giardia, and analysis should be performed by a laboratory which has been accredited under international standard ISO/IEC 17025 (ISO/IEC 2005). In South Africa, the South African National Accreditation System (SANAS) is the body responsible for accreditation of analytical laboratories (Du Preez, 2000).

STEP VII: Risk categorisation Very High Risk
Risk Score > 100

High Risk
Risk Score 50 -99

Low Risk
Risk Score < 10

Figure 18
Characterisation of risk using the risk assessment scores  (2003) be followed for sample processing. Quality control procedures must be carefully followed and adhered to in order to render the results credible.

Application
The sample collection and analyses were conducted by Rand Water Analytical Services Laboratory which is accredited under International Standard ISO/IEC 17025 (ISO/IEC, 2005). A 10 ℓ volume of water was collected in a carboy and transported to the laboratory. The sample was analysed using Rand Water Method 06 which involves sample filtration, elution, separation and staining. Quality was assured through reproducible calibration and testing of the filtration, immunomagnetic  (RW, 2010). Qualitative data analysis was performed by scanning each slide well for objects that meet the size, shape, and fluorescence characteristics of Cryptosporidium oocysts or Giardia cysts. Potential oocysts or cysts were confirmed through DAPI staining characteristics and DIC microscopy. Quantitative analysis was performed by counting the total number of objects on the slide confirmed as oocysts or cysts.
Step IX: Data evaluation, interpretation and storage (Fig. 20) The results of the analysis were reported as oocysts or cysts per 10 ℓ. Reporting of non-compliant data must be done according to the available protocols. In South Africa, data is evaluated against the National Drinking Water Quality Standard (SANS 241-2011). Any incidence of Cryptosporidium or Giardia in treated water at a concentration of 1 or more oocysts in 10 ℓ is dealt with following the procedures described in the Protozoan Incident Management Framework of South Africa as set out in DWAF (2007). All of the results should be stored in the database provided by the company. At Rand Water, for example, data is stored in the laboratory information management system (LIMS). Alternatively, data can be stored at the facilities of the governing authority. An example of such an authority in South Africa

Figure 19
Activities 310 is the Department of Water Affairs (DWA) which is the custodian of drinking water in South Africa. Furthermore data reports should be compiled and these should be stored for at least 3 years.

Application
The results of the analysis were reported as oocysts or cysts per 10 ℓ. The results were stored in the LIMS database. Data reports were compiled and these are stored for at least 3 years (RW, 2010). These are reported to DWA as a part of Blue Drop requirements.

PHASE IV: Monitoring process review
During this phase, the date of the next review will be determined.
Step X: Process evaluation and review (Fig. 21) During the review the mitigation measures taken should be considered. These should be noted in the desktop survey. The review should take place every 3 years or if there is significant deterioration of the source water quality, a suspected cryptosporidiosis or giardiasis outbreak or in the event of plant disturbance which could possibly result in the leakage of Cryptosporidium and/or Giardia oocysts into the treated water (NI EHS, 2002). However, in the case of very high risk areas the frequency of monitoring could be increased and the mitigation measures be imposed until the results are satisfactory. An example of such measures is the analysis of water samples following the incident management protocol set out in DWAF (2007). The desktop survey should be re-assessed in order to ensure that there are no major changes in the regulations.

Application
Evaluation of the Vaal River Barrage catchment indicated that it is a very high risk area for Cryptosporidium and Giardia. Samples should be collected at the points of the water treatment system identified as very high risk. The sampling frequency for these areas should be once a week at the source water, twice a week at the abstraction point and 3 times a week on the final water until the risk has been reduced to

311
'medium' or 'minimum'. However, if Cryptosporidium oocysts and/or Giardia cysts are found in drinking water samples, a corresponding sampling frequency should accordingly be adopted, as stipulated in the Protozoan Incident Management Framework. This involves increasing the sampling frequency, notification of the municipality, Department of Water Affairs and Department of Health (DWAF, 2007). In cases where there is an expansion in the water supply or the capacity of the water treatment plant is exceeded and if there are any operational changes in the plant, the entire water supply system will be evaluated for risk using this protocol. During the review the mitigation measures taken should be considered and these should be noted in the desktop survey. The review will take place every 3 years or if there is significant deterioration of the raw water quality, if there is a suspected cryptosporidiosis or giardiasis outbreak or in the event of plant disturbance which could possibly result in the leakage of Cryptosporidium and/or Giardia oocysts into the treated water (NI EHS, 2002). However, in case of high risk areas the frequency of monitoring could be increased and the mitigation measures be imposed until the results are satisfactory.

CONCLUSIONS
The primary objective of the study was to develop a protocol/ methodology that can be applied by drinking water producers to monitor Cryptosporidium and Giardia. The developed protocol was applied at a drinking water plant receiving source water from the Vaal River Barrage reservoir. Applying the protocol showed that it was easy to follow and resulted in the identification of the risk areas and potential risks in the water supply system. The approach of developing a protocol and testing it under operational conditions appeared to be effective and should be followed in future studies. The only constraint is that the study was tested on one water treatment plant. From the study, it can be concluded that the source water from the Vaal River Barrage catchment was a very high risk. However, the use of the multibarrier approach coupled with advanced treatment using UV rendered the water almost free of Cryptosporidium and Giardia. Therefore, the population receiving treated water from the Barrage purification plant is not at risk of cryptosporidiosis and giardiasis.
Using the protocol could also benefit water utilities as it can assist managers to evaluate their source water catchments and establish the Cryptosporidium and Giardia risk status of the water resource. Improvement strategies, for example, negotiations with the local and national government to improve catchment management strategies, can thus be implemented. The water utility can decide to reduce the risk of Crypto sporidium and Giardia contamination by implementing advanced treatment processes or by upgrading their treatment processes, particularly through the addition of UV or ozone treatment. In addition, the protocol can be used by water utilities as part of their monitoring strategy as required in Water Safety Planning.
The approach can also be used by the catchment management agencies as one of the tools that will assist them in identifying the potential risks that could have an impact on the catchments. The developed protocol is thus beneficial to governmental agencies as well as water utilities tasked to ensure the production of safe drinking water for consumers.